# Lectures, seminars and dissertations

* Dates within the next 7 days are marked by a star.

Emil Sköldberg

**The homologies of monomial ideals and incidence algebras.**

*** ** Wednesday 25 October 2017, 15:15, M2 (M233)

Given a set of monomials M ={m_1, ..., m_r} in a polynomial ring,
one can form the set L of all least common multiples of
subsets of M. This set is a partially ordered set, (even a lattice)
ordered by divisibility.
It is known that the betti numbers of the ring S/(M) can be computed
as the homology of the simplicial complex of L. In the talk,
I will explain how this relationship can be lifted to hold on the level
of resolutions, between modules over S on one hand, and modules over
the incidence algebra of L on the other hand.

ANTA seminar

Prof. Roland Griesmaier (University of Wuerzburg)

**Uncertainty principles for far field patterns and applications to inverse source problems**

*** ** Thursday 26 October 2017, 12:30, M2 (M233)

Seminar on Applied Mathematics

Dissertation

Lauri Mustonen, Prof. Roland Griesmaier (Wuerzburg)

**Approximations and surrogates for computational inverse boundary value problems**

*** ** Friday 27 October 2017, 12:00, M1 (M232)

Prof. Lothar Nannen (TU Wien)

**Numerical methods for resonance problems in open systems**

Tuesday 31 October 2017, 15:00, U1

Colloquim

David Radnell (Aalto University)

**Model of the Teichmuller space of genus-zero bordered surfaces by period maps**

Wednesday 01 November 2017, 12:15, M3 (M234)

The classical period mapping takes compact Riemann surfaces of genus g into the Siegel upper half-plane, which consists of symmetric g x g matrices with positive definite imaginary part. It is a classical fact that this map is holomorphic.
We generalize the period mapping to the case of the Teichmuller space of genus-zero surfaces with n closed non-overlapping disks removed and prove it is holomorphic. The period mapping takes the Teichmuller space of this type into the direct product of the Teichmuller space of genus-zero surfaces with n punctures and a space of bounded operators on an n-fold sum of Bergman spaces of the disk. The construction relies on a generalization of the classical Grunsky operator, new extension and reflection results on quasidisks, as well as a coordinate system for the infinite-dimensional Teichmuller space inspired by conformal field theory.
This talk will introduce these results which are joint work with Eric Schippers and Wolfgang Staubach. Background notions will be explained and technical details will be hidden.

Seminar on analysis and geometry

Laura Venieri (University of Helsinki)

**TBA**

Wednesday 08 November 2017, 12:15, M3 (M234)

Seminar on analysis and geometry

Matias Vestberg (Aalto)

**TBA**

Friday 10 November 2017, 14:15, Kumpula Exactum C124

Harmonic analysis and PDE seminar

Alessandro Mancuso (Aalto University)

**Bayesian approach for safety barrier optimization**

Tuesday 14 November 2017, 11:15, M205

Alessandro Mancuso

**TBA**

Tuesday 14 November 2017, 11:15, M205

Doctoral student talk / Engström

M.Sc. Edoardo Tosoni

**Safety assessment of nuclear waste repositories**

Tuesday 14 November 2017, 11:30, M205

Edoardo Tosoni

**TBA**

Tuesday 14 November 2017, 11:40, M205

Doctoral student talk / Engström

Vilma Virasjoki

**Decision Models of Energy Markets for Investment and Policy Analysis**

Tuesday 14 November 2017, 12:15, M205

Doctoral student talk / Engström

Lauri Hitruhin (University of Helsinki)

**TBA**

Wednesday 15 November 2017, 12:15, M3 (M234)

Seminar on analysis and geometry

Prof. Kari Astala (Helsinki University, Aalto University)

**TBA**

Tuesday 28 November 2017, 15:00, U1

Colloquim

Matias Heikkilä

**On multivariate separating Hill estimator under estimated location and scatter**

Tuesday 12 December 2017, 11:15, M205

Joona Karjalainen

**Moment-based parameter estimation in random intersection graphs**

Tuesday 12 December 2017, 11:40, M205

Doctoral student talk / Engström

Alex Karrila

**Uniform spanning trees, scaling limits, and conformal invariance**

Tuesday 12 December 2017, 12:15, M205

Doctoral student talk / Engström

Jouni Parkkonen (University of Jyväskylä)

**TBA**

Wednesday 24 January 2018, 12:15, M3 (M234)

Seminar on analysis and geometry

## Past events

Jia Huilian (Aalto Xián Jiaotong University)

**Global regularity of elliptic p-Laplace equation on convex domain**

Friday 20 October 2017, 15:15, M3 (M234)

Harmonic analysis and PDE seminar

Tuomas Hytönen (Helsinki)

**The matrix-weighted frontier of sharp norm inequalities**

Friday 20 October 2017, 14:15, M3 (M234)

Harmonic analysis and PDE seminar

Stavros Evdoridis

**Bohr's inequality for harmonic mappings**

Wednesday 18 October 2017, 12:15, M3 (M234)

Seminar on analysis and geometry

Robin Rajamäki

**Additive 2D bases**

Wednesday 11 October 2017, 15:15, M2 (M233)

Additive bases, i.e. sets of integers whose pairwise sums cover a desired set of integers, have been studied by number theorists since the early 20th century. In particular, the combinatorial problem of minimizing the number of elements required to cover a given set of consecutive integers is often of interest. Such optimal bases find applications in e.g. sparse linear sensor arrays, where the number of sensors is desired to be kept low in order to reduce costs and mitigate non-ideal coupling effects between array elements. Although much of the existing work on additive bases is directly applicable to 1D sparse array design, often 2D arrays are required in practice. However, additive 2D bases have not been studied as extensively as their 1D counterpart in the past. In our talk, we present some recent results on rectangular additive bases, i.e. additive bases whose sumsets cover a rectangular area of consecutive points in the plane. For example, optimal rectangular bases found through exhaustive computer search are shown. Furthermore, different rectangular bases with analytically tractable structure and a low element count are introduced.

ANTA seminar

Galia Dafni (Concordia University)

**Local BMO and Van Schaftingen spaces**

Wednesday 11 October 2017, 12:05, M3 (M234)

Abstract: Based on an inequality of Bourgain and Brezis, Van Schaftingen introduced a scale of function spaces lying between the critical Sobolev space and BMO, and suggested that an analogous result should hold in the case of local BMO spaces on domains. We discuss recent work of A. Butaev, who constructed nonhomogeneous versions of Van Schaftingen spaces on Euclidean space, as well as on manifolds, and showed that they lie between the critical Sobolev space and the space bmo, defined by Goldberg. Moreover, he proved a corresponding version for the nonhomogeneous space of functions of vanishing mean oscillation (vmo). On a Lipschitz domain $\Omega$, Butaev answered the question of Van Schaftingen by showing that there are two scales of spaces lying between $W^{1,n}(\Omega)$ and $W^{1,n}_0(\Omega)$, respectively, and the local BMO
spaces bmo_r and bmo_z studied by Chang-Krantz-Stein and Chang-Dafni-Stein.

Seminar on analysis and geometry

Niko Väisänen (Nokia/Aalto)

**Beamforming -- Enabling Higher Data Rate Wireless Communications**

Thursday 05 October 2017, 15:15, M205

MSc thesis presentation

ANTA

Kristian Moring (Diploma thesis talk)

**Direct methods in the calculus of variations for parabolic problems**

Thursday 05 October 2017, 15:15, M3 (M234)

Mateus Sousa (IMPA, Rio de Janeiro)

** Existence for Fourier restriction on hyperboloids**

Thursday 05 October 2017, 14:15, M3 (M234)

In this talk we present some results about existence, as well as non existence, of extremizers for the adjoint Fourier restriction estimate on hyperboloids. We will discuss the Lorentz symmetries of the problem and the concentration-compactness arguments involved in the proofs, as well as connections to the Klein-Gordon equation.

Negin Karimi

**LCD codes**

Wednesday 04 October 2017, 15:15, M3 (M234)

Error-correcting codes play an important role in digital communication among all types of codes. Linear codes are studied the most. Because of their algebraic structure, they are easier to describe, encode, and decode than nonlinear codes.
A special class of linear codes is linear complementary dual code (LCD code). LCD codes are defined by Massy in 1992. He was shown that asymptotically good codes exist. LCD codes have been widely applied in data storage, communications systems, and cryptography. Also, they are interesting objects in the general framework of algebraic coding. In this talk will be presented some of the properties cyclic codes, quasi-cyclic codes and quasi-twisted codes with complementary dual.

ANTA seminar

Juha Videman (University of Lisbon)

**A degenerate elliptic-parabolic system arising in competitive contaminant transport **

Wednesday 04 October 2017, 12:15, M3 (M234)

The transport of contaminants in subsurface environments is a complex process modeled by
advection-diffusion-reaction equations. If the reactive adsorption process (accumulation of a
pollutant on the porous matrix) is described by a nonlinear isotherm of Freundlich type, the
evolution of the solute concentrations is governed by a degenerate parabolic equation. In the
multi-species case (contamination by several solutes), this corresponds to a coupled system of
degenerate elliptic-parabolic equations.
After introducing the basic concepts of modeling contaminant transport in porous media, I will
show that this system admits a unique weak solution provided the nonlinear adsorption isotherm
associated with the reaction process satisfies certain physically reasonable structural conditions.
Under these conditions, the system falls under the general quasi-linear elliptic-parabolic systems
addressed by Alt and Luckhaus in their seminal work [1]. The main difference between our
approach and that of [1] lies in the existence proof which we base on Rothes method and on
solving a convex minimization problem at each time step. Our approach provides also a simple
method for the numerical approximation of the system.
This is a joint work with Margarida Baia (CAMGSD/IST), Farid Bozorgnia (CAMGSD) and
Leonard Monsaingeon (CAMGSD/Universite de Lorraine).
[1] W. H. Alt and S. Luckhaus, Quasilinear Elliptic-Parabolic Dierential Equations, Math. Z.
183 (1983), 311-341.

Seminar on analysis and geometry

Timo Hänninen (HY)

**Two-weight Lp-Lq bounds for positive dyadic operators in the case 0**

Friday 29 September 2017, 15:15, Kumpula Exactum C124

Harmonic analysis and PDE seminar

**Thomas Singer (Aalto)**

**Existence of variational solutions in non-cylindrical domains**

Friday 29 September 2017, 14:15, Kumpula Exactum C124

Harmonic analysis and PDE seminar

**Esko Heinonen (University of Helsinki)**

**Existence and non-existence results for minimal graphic and p-harmonic functions**

Wednesday 27 September 2017, 12:15, M3 (M234)

In the Euclidean space, by the celebrated result due to Bombieri, De Giorgi, and Miranda, all positive entire solutions of the minimal graph equation are constant. It turns out that on Riemannian manifolds similar results can be obtained for solutions with at most linear growth if the manifold has only one end and asymptotically non-negative sectional curvature. In this talk I will discuss about recent results concerning the existence and non-existence of entire minimal graphic and p-harmonic functions. Talk is based on joint work with Jean-Baptiste Casteras and Ilkka Holopainen.

Seminar on analysis and geometry

** Prof. Daniele Boffi (Università di Pavia, Aalto University) **

**Finite element approximation of resonant modes for the Maxwell cavity problem**

Tuesday 26 September 2017, 15:15, M1 (M232)

The resonant modes of time harmonic Maxwell equations can be identified with the eigensolutions of a problem represented by a variational formulation in mixed form. In this talk we review the finite element approximation of eigenvalue problems arising from partial differential equations. In particular, it will be shown that the analysis of mixed problems differs significantly from that of standard Galerkin formulations. We will show how these results can be applied to the a priori analysis of Maxwell's eigenvalue problem and we will give some indications on how the a posteriori analysis can be performed

Colloquim

**Lauri Harhanen (KaVo Kerr), Antti Huhtala (SSF), Saara Hyvönen (DAIN Studios), Mika Juntunen (KONE), Juho Könnö (Wärtsilä), Otto Seiskari (IndoorAtlas), Stratos Staboulis (Eniram)**

**Sovelletut matemaatikot teollisuudessa**

Thursday 21 September 2017, 14:00, U6, Otakaari 1

** Jeff Lindquist (University of California, LA)**

**Weak capacity in Ahlfors regular metric spaces**

Wednesday 20 September 2017, 12:15, M3 (M234)

We construct and use a hyperbolic filling of a $Q$-regular
compact doubling metric space $Z$ to define the notion of the weak
$p$-capacity between appropriate subsets of $Z$. This notion extends
modulus and is preserved up to constants by quasisymmetric maps. We
explore some applications involving Ahlfors regular conformal dimension
and quasisymmetric uniformization of metric 2-spheres.

Seminar on analysis and geometry

**Tuomo Tiilikainen**

**Lämpöyhtälön keskiarvoperiaate/Diplomityöesitelmä**

Wednesday 13 September 2017, 15:15, M2 (M233)

**Ilmari Kangasniemi (University of Helsinki)**

**Cohomological behavior of uniformly quasiregular mappings**

Wednesday 13 September 2017, 12:15, M3 (M234)

Based on a joint work with Pekka Pankka. Given a uniformly quasiregular map f on a closed oriented Riemannian manifold, it turns out that all the complex eigenvalues of the induced pull-back map f* in real singular cohomology have a modulus of (deg f)^(k/n), and moreover the induced cohomological map is complex diagonalizable. This can be exploited to obtain limitations on the possible degrees of uniformly quasiregular maps on closed manifolds. The talk goes over some basics and context regarding uniformly quasiregular mappings, and presents the main ideas behind the proofs of the aforementioned results.

Seminar on analysis and geometry

**D.Sc (Tech) Antti Punkka**

**Teaching demonstration: Pareto-optimality in solving multicriteria optimization problems**

Friday 08 September 2017, 10:30, M2 (M233)

**Jan Hämäläinen**

**Elementtimenetelmä magnetostatiikassa**

Thursday 07 September 2017, 11:15, M237

Kandiseminaari

**Tarmo Kivioja**

**Lokaalin sileyden estimoinnista**

Thursday 07 September 2017, 10:15, M237

Kandiseminaari

**Ville Kujala**

**Achieving the capacity of coded PIR using arbitrary linear code (BSc presentation)**

Friday 25 August 2017, 15:15, M3 (M234)

Kandidaattiseminaari

**Juuso Korvuo**

**Hilojen lähimmän vektorin ongelma (kandiesitelmä)**

Thursday 24 August 2017, 14:15, M3 (M234)

Kandidaattiseminaari

**Aki Malinen**

**Algebraic methods in maximum likelihood estimation (BSc presentation)**

Thursday 24 August 2017, 09:15, M3 (M234)

Kandidaattiseminaari

**Hoa Ngo [Stochastic afternoon]**

**Information spreading in a large network**

Tuesday 22 August 2017, 16:10, M3 (M234)

A simple mathematical model for information spreading is a complete graph where everyone knows everyone and everyone relays messages at random time instants to randomly chosen neighbours independently of each other.
In this talk we will consider information spreading in the large configuration model, which is a more advanced model. One of the key quantity analysing the rumour spreading speed is the broadcast time, the time when the rumour has reached the entire population. We will also discuss the broadcast time taking into account the passive and active users.

**Alex Karrila [Stochastic afternoon]**

**Boundary branches in a uniform random spanning tree of a planar graph**

Tuesday 22 August 2017, 15:10, M3 (M234)

A physics principle asserts that the scaling limit of a critical lattice model, as increasingly dense lattices approximate a continuum domain, is described by a conformal field theory (CFT). The aim of this talk is to prove rigorously conformal invariance properties in the scaling limit of the uniform spanning tree (UST) of a planar graph, i.e., a tree subgraph that covers all the vertices of the original graph, chosen uniformly at random. In a spanning tree, any two vertices are connected by a unique path, called a branch. We study multiple simultaneous UST boundary-to-boundary branches between given boundary vertices, as well as the boundary visits of a single such branch. The related probabilities have conformally invariant scaling limits, and solve partial differential equations as predicted by CFT. As a collection of curves, such multiple simultaneous branches converge weakly to a conformally invariant law, called the local multiple SLE(2). These are among the first verifications of third-order PDEs of CFT, as well as convergence results to multiple SLE.

**Armando W. Gutiérrez [Stochastic afternoon]**

**The horofunction compactificacion of lp spaces and Hilbert's projective metric**

Tuesday 22 August 2017, 14:00, M3 (M234)

The horofunction compactification is the result of making any metric space into a compact topological space
by only using the metric. The elements of this compactification have been recently shown to be very useful in the study
of limit theorems for deterministic and random dynamical systems.
In this talk I will give a complete description of the horofunction compactification of the classical lp spaces.
I will also consider Hilbert's projective metric on the standard cone of positive real vectors, and describe its horofunction compactification. The latter will be used to give a new proof of the well-known Perron theorem.

**Valentina Candiani (University of Genova)**

**The role of segmentation in kidney compartmental analysis**

Tuesday 22 August 2017, 10:15, M2 (M233)

Seminar on Applied Mathematics

**Prof. Saminathan Ponnusamy (ISI Chennai/IIT Madras)**

**Bohr radius**

Wednesday 16 August 2017, 12:15, M3 (M234)

Seminar on Analysis and Geometry

**Mikko Kivelä (Aalto University)**

**Randomized reference models and spreading in temporal networks**

Monday 14 August 2017, 15:15, M3 (M234)

For a long time network science concentrated on static graphs as representations of networked systems. This abstraction was used to analyse shortest path lengths, spreading of disease or information in networks, and many other things. Often the underlying assumption behind such analysis was that the activation of nodes and links is controlled by homogeneous Poisson processes. More recently there has been growing interest in analysing 'temporal networks' where the link activation times are determined directly from data. In this talk I will illustrate how this approach can be used to analyse a large communication network with hundreds of millions of link activation events. I will focus on how 'reference models', in which the data is shuffled in various ways, can be used for this data and how they are used in the literature on temporal networks in general. I will also discuss the challenges in the literature that are caused by the sudden increase in use of such shuffling methods of temporal network.

Aalto Stochastics & Statistics Seminar

**Kaisa Nyberg / Chris Brzuska**

**Connections between linear and statistical independence of binary random variables / Survey about indistinguishability obfuscation **

Monday 14 August 2017, 14:30, M2 (M233)

Informal presentations during the visit of our new Cryptology professor Chris Brzuska

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Dissertation**

Vesa Kaarnioja, Prof. Lars Grasedyck (RWTH Aachen University)

**On sparse tensor structures in lattice theory and applications of the polynomial collocation method based on sparse grids**

Wednesday 02 August 2017, 12:00, M1 (M232)

Further information

**Leevi Korkeala**

**Hyperbolisten ryhmien pinta-aliryhmät (kandiesitelmä)**

Wednesday 02 August 2017, 11:15, M3 (M234)

Kandidaattiseminaari

**Miio Taarna **

**Julkisen liikenteen verkostojen tarkastelu verkkoteorian keinoin (kandiesitelmä)**

Wednesday 02 August 2017, 10:15, M3 (M234)

Kandidaattiseminaari

**Pauliina Kärkkäinen**

**Positroniemissiotomografian käänteisongelma**

Wednesday 02 August 2017, 09:15, M3 (M234)

Kandidaattiseminaari

**Valtteri Lipiäinen**

**Verkkojen Ollivier-Ricci -kaarevuus (kandiesitelmä)**

Thursday 20 July 2017, 11:15, M3 (M234)

Kandidaattiseminaari

**Miika Leinonen**

**Kvaternioalgebrat ja hyvin pyöristyvät hilat (kandiesitelmä)**

Thursday 20 July 2017, 10:15, M3 (M234)

Kandidaattiseminaari

**Prof. Gonzalo R. Arce (University of Delaware)**

**Coded Aperture Optimization in Compressive X-ray Tomography **

Thursday 06 July 2017, 13:15, M205

Seminar on Applied Mathematics

**Prof. Oktay Olmez, Ankara University**

**Binary three-weight linear codes from partial geometric difference sets**

Wednesday 21 June 2017, 15:00, M3 (M234)

Links between linear codes, non-linear functions from cryptography, graphs and combinatorial designs have attracted the attention of many researchers over the last 50 years. Difference set method is a powerful tool to construct designs and explore the links between designs and many other combinatorial objects including codes and nonlinear cryptographic functions.
In this talk, we will introduce a generalisation of $(v,k,\lambda)-$difference sets known as partial geometric difference sets. In particular, we will show that existence of a family of partial geometric difference sets is equivalent to existence of a certain family of three-weight linear codes. We also provide a link between binary plateaued functions, three-weight linear codes and partial geometric difference sets.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Kurkela, Korkeala, Stenbacka, Wilkman, Hämäläinen ja Taarna**

**Kanditöiden aihe-esittelyt, sessio 3**

Monday 19 June 2017, 15:00, M3 (M234)

Sessiossa kandityönsä aiheen esittelevät Lauri Kurkela, Leevi Korkeala, Filip Stenbacka, Markus Wilkman, Jan Hämäläinen ja Miio Taarna

Hollanti & Kangaslampi

**Kiuru, Kivioja, Kujala, Malinen, Hyytiäinen ja Koskinen**

**Kanditöiden aihe-esittelyt, sessio 2**

Monday 19 June 2017, 14:00, M3 (M234)

Sessiossa kandityönsä aiheen esittelevät Aukusti Kiuru, Tarmo Kivioja, Ville Kujala, Aki Malinen, Tatu Hyytiäinen ja Tuomas Koskinen

Hollanti & Kangaslampi

**Kärkkäinen, Lipiäinen, Tuomala, Aaltonen, Korvuo ja Leinonen**

**Kanditöiden aihe-esittelyt, sessio 1**

Monday 19 June 2017, 13:15, M3 (M234)

Sessiossa kandityönsä aiheen esittelevät Pauliina Kärkkäinen, Valtteri Lipiäinen, Lari-Matti Tuomala, Tuuli Aaltonen, Juuso Korvuo ja Miika Leinonen

Hollanti & Kangaslampi

**Christopher Hopper **

**Fractional differentiability and singular set bounds for holonomic minimisers of integral functionals**

Friday 09 June 2017, 14:15, M3 (M234)

Harmonic analysis and PDE seminar

**Dr. Fabricio Oliveira (RMIT Australia)**

**An introduction to Stochastic programming: modelling and solving decision problems under uncertainty**

Wednesday 07 June 2017, 11:00, M2 (M233)

Teaching demonstration

**Dr. Fabricio Oliveira (RMIT Australia)**

**Efficiently solving stochastic mixed-integer problems combining Gauss-Seidel and penalty-based methods**

Wednesday 07 June 2017, 09:00, M2 (M233)

**Cordian Riener (Kontanz + Arctic University)**

**Semidefinite programming and arithmetic progressions**

Thursday 01 June 2017, 16:45, M240

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**Louis Theran (St Andrews)**

**Generic global and universal rigidity and the power of SDP for graph realisation**

Thursday 01 June 2017, 16:00, M240

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**Parinya Chalermsook (Aalto CS)**

**Towards Fine-grained Combinatorial Optimization for NP-hard Problems**

Thursday 01 June 2017, 14:45, M240

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**Camilla Hollanti (Aalto MATH)**

**Private information retrieval from coded databases**

Thursday 01 June 2017, 14:00, M240

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**Aristides Gionis (Aalto CS)**

**Aligning networks: many and actively**

Thursday 01 June 2017, 13:00, M240

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**Chris Brzuska**

**Foundations of cryptography and their applications**

Tuesday 30 May 2017, 13:30, T5, CS building

**Chris Brzuska**

**Meet-in-the-middle attack **

Tuesday 30 May 2017, 09:00, M1 (M232)

**Nico Döttling**

**Powerful Primitives for Weak Assumptions**

Monday 29 May 2017, 13:30, T2, CS building

**Nico Döttling**

**Meet-in-the-middle attack **

Monday 29 May 2017, 09:00, M1 (M232)

**Arnab Roy**

**Secure and Efficient Symmetric-key Primities for Multiparty Computation/Zero-Knowledge/Verifiable Computational **

Wednesday 24 May 2017, 13:30, T5, CS building

**Smiljana Jaksic (University of Belgrade, Serbia)**

**Spaces of ultradistributions with applications to pseudo-differential operators**

Wednesday 24 May 2017, 12:15, M3 (M234)

This talk is based on joint works with Stevan Pilipovic and Bojan Prangoski.
The first part of the talk is devoted to spaces of test functions and ultradistributions on the positive orthant of the Euclidean space,
which can be described as analogous to the Gelfand-Shilov spaces and their dual spaces.
Such ultradistributions are characterized through the Laguerre expansions.
Furthermore, the full topological description of the spaces is given.
The second part is devoted to the class of the Weyl pseudo-differential operators
with radial symbols from the ultradistribution spaces.
The continuity properties of these classes of pseudo-differential operators
over the Gelfand-Shilov spaces and their dual spaces are proved.
In this way, the classes of the Weyl pseudo-differential operators are extended
to those with the radial symbols with the exponential and sub-exponential growth rate.

Seminar on analysis and geometry

**Kai Virtanen**

**Solving Dynamic Optimization Problems**

Wednesday 24 May 2017, 10:30, M2 (M233)

Teaching demonstration

**Arnab Roy**

**Meet-in-the-middle attack **

Wednesday 24 May 2017, 09:00, M1 (M232)

**Prof. Samuli Siltanen (University of Helsinki)**

**Electrical impedance tomography imaging of stroke**

Tuesday 23 May 2017, 14:15, M3 (M234)

Seminar on Applied Mathematics

**Vincent Grosso**

**Cryptographic implementations and physical defaults**

Tuesday 23 May 2017, 13:30, T2, CS building

**Vincent Grosso**

**Meet-in-the-middle attack **

Tuesday 23 May 2017, 09:00, M1 (M232)

**Esa Hyytiä (University of Iceland)**

**TBA**

Monday 22 May 2017, 17:30, M1 (M232)

**Sergei Chubanov (University of Siegen)**

**TBA**

Monday 22 May 2017, 16:45, M1 (M232)

**Pierre Nyquist (KTH)**

**TBA**

Monday 22 May 2017, 16:00, M1 (M232)

**Sergei Chubanov (University of Siegen)**

**Optimality conditions for unconstrained convex optimization**

Monday 22 May 2017, 10:00, M1 (M232)

**Esa Hyytiä (University of Iceland)**

**Routing jobs to servers**

Monday 22 May 2017, 09:30, M1 (M232)

**Pierre Nyquist (KTH)**

**Time series analysis, specifically (G)ARCH processes for financial time series**

Monday 22 May 2017, 09:00, M1 (M232)

**Dimitrios Ntalampekos (UCLA)**

**Uniformization of Sierpinski carpets by square carpets**

Friday 19 May 2017, 14:15, M3 (M234)

Uniformization of metric spaces is the problem of finding conditions on a metric space, under which it can be transformed to a canonical space, by a map that preserves the geometry. In this talk, our metric space will be planar Sierpinski carpets, and the canonical spaces are square Sierpinksi carpets. We prove that every Sierpinski carpet, under certain geometric assumptions, can be mapped by a quasisymmetric map to a square carpet. This is achieved with the aid of carpet-harmonic functions, which is a discrete notion of harmonic functions on Sierpinski carpets.

Harmonic analysis and PDE seminar

**Prof. Daina Taimina (Cornell University)**

** Study mathematics and ... become an artist?**

Tuesday 16 May 2017, 15:00, C-hall (Y205)

When I was in middle school I was a good student and had very good grades in all subjects...except in art class where I was struggling to draw the way my teacher thought was the right way. I accepted my teacher's judgement that I can do anything but art and went to study mathematics. Could I imagine that some fifty years later I will be known as an artist? Mathematics and art are not too distant. In my talk, I will discuss the interplay between mathematics and art from my own perspective.

Joint Mathematics and Arts and Women in Mathematics Colloquium / Kirsi Peltonen

**Céline Blondeau**

**Statistical attacks on symmetric cryptographic primitives (Job talk for Tenure Track in Cryptology position)**

Friday 12 May 2017, 14:00, T2, CS building

**Céline Blondeau**

**Meet-in-the-middle attack (Teaching demonstration for Tenure Track in Cryptology position)**

Friday 12 May 2017, 09:00, M1 (M232)

**Dissertation**

Amaro Barreal

**Doctoral thesis defense: Lattice codes for physical layer communications**

Thursday 11 May 2017, 14:00, M1 (M232)

Opponent Prof. Bharath Sethuraman, custos Camilla Hollanti.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Thomas Westerbäck**

**Combinatorial coding theory via polymatroids**

Wednesday 10 May 2017, 16:15, M3 (M234)

Matroid theory can be used to analyze many interesting properties of linear codes over finite fields. Recent research has proven matroid theory to be a valuable tool in several areas in coding theory, e.g. in distributed storage, index coding and network coding. Polymatroids, a generalization of matroids, can be associated with any finite code. In this talk I will present how polymatroids and codes are closely related via entropy and how polymatroid theory can be used in order to analyze many interesting properties of codes. New coding theoretical results will be given in the setting of polymatroids. These results can therefore also be applied to non-code objects that are associated with polymatroids.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Prof. Bharath Sethuraman (California State University, Northridge)**

**Matrices from Division Algebras and Fast Decodability (NOTE: ON TUESDAY!)**

Tuesday 09 May 2017, 15:15, M2 (M233)

We call two matrices A and B in M_n(C) mutually
orthogonal if AB* + BA*=0, where A* indicates the conjugate
transpose of A. The situation where such matrices arise from the
embedding of a division algebra over a number field into
M_n(C) is particularly interesting in wireless communications,
where the presence of such matrices leads to fast decodability. We
describe fundamental limits on the number of families of matrices that
are mutually orthogonal across the families, and limits on the number of
matrices in each family. The limits are obtained by a mix of techniques
from both linear algebra and the theory of Brauer groups of commutative
rings.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Robert Krone (Queen's University)**

**Rigid boundary components of non-negative rank matrices**

Tuesday 09 May 2017, 10:15, M237

The non-negative rank of a matrix is useful for algebraic statistics, but difficult to compute. The set of matrices with rank and non-negative rank both equal to r forms a semi-algebraic, Zariski-dense subset of the variety of rank r matrices. Describing all the boundary components of this set can provide a membership test, and this was accomplished for rank 3 by Mond, Smith and van Straten. Using a slightly different approach, we describe some of the boundary components in the higher rank cases, particularly the set of matrices that have a rigid nonnegative decomposition.

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**President Tuula Teeri, CEO Tapio Koivu (Heureka)**

**Sensual Mathematics**

Monday 08 May 2017, 18:00, Heureka

Further information

Opening of the Exhibition Sensual Mathematics at Heureka

Kirsi Peltonen

**Prof. David Rios Insua (ICMAT-CSIC and Royal Academy of Sciences, Spain)**

**Adversarial Risk Analysis: Concepts, Applications and Challenges**

Tuesday 02 May 2017, 15:15, M1 (M232)

Our focus will be on problems in which two or more agents confront (as in non-cooperative game theory) and the consequences that they receive are random (as in risk analysis) and depend on the actions of all participants (as in game theory). This is motivated by problems in relation with high-profile terrorist attacks, business decision making and cyber-security.
Most authors in this context would use game theoretical methods based on variants of Nash equilibria concepts. However, this is not always satisfactory in most of the above applications since beliefs and preferences of adversaries will not be readily available, frequently violating game theoretic common knowledge assumptions. In contrast, we focus on Adversarial Risk Analysis (ARA), an emergent paradigm for the type of problems we consider. ARA provides one-sided prescriptive support to a DM, maximizing her subjective expected utility, treating the adversaries' decisions as random variables. ARA models the adversaries' decision-making problems and, under assumptions about their rationality, tries to assess their beliefs and preferences. Then, ARA can predict their optimal actions. However, the uncertainty about the adversaries' judgements is propagated into their decisions, leading to random optimal adversarial decisions which provide the necessary distributions over the adversaries' decision. The main goals of ARA are to weaken standard common knowledge assumptions and provide more flexible models for opponent behavior. ARA is explicitly Bayesian in that subjective distributions are employed to express the uncertainties of the analyst.
We shall outline basic concepts in ARA, some relevant applications and identify a few challenges.

Colloquium of Mathematics and Systems Analysis

**Pekka Lehtelä**

**On the regularity of supersolutions to the porous medium equation**

Friday 28 April 2017, 14:15, M3 (M234)

Harmonic analysis and PDE seminar

**Tommi Brander, University of Jyväskylä**

**Detecting inclusions in Calderón's problem for p-Laplacian**

Thursday 27 April 2017, 14:15, M3 (M234)

Seminar in Applied Mathematics

**Dr. David Cushing (Durham)**

**Bakry-Emery curvature functions of graphs**

Wednesday 26 April 2017, 15:15, M3 (M234)

Ricci curvature has proven to be a vital notion in many areas of mathematics. There have been various attempts at generalising this notion to graphs. We look at one such notion due to Bakry and Emery. We introduce the Bakry-Emery curvature function and obtain results on Cartesian products of graphs and also on a certain regularity property. I will also demonstrate interactive software that computes curvature.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Prof. Harri Ehtamo (Aalto)**

**Evoluutiopelit evakuointimalleissa - Pelastettavien kovan leikin peliteoreettinen kuvaus**

Wednesday 26 April 2017, 15:00, U8

LUMA-luennot lukiolaisille / Kangaslampi

** Stavros Evdoridis (Aalto)**

**On Geometric Properties of Polyharmonic Mappings**

Wednesday 26 April 2017, 12:15, M3 (M234)

Seminar on analysis and geometry

**Thomas Britz (UNSW Sydney)**

**A Nice Proof of Wei's Duality Theorem**

Tuesday 25 April 2017, 15:15, M1 (M232)

The Tutte polynomial was once an esoteric object known only to the then small community of combinatorialists. That changed when Greene (1976) pointed out the connection between this polynomial and weight enumerators, and how that connection provided a beautifully simple proof of the MacWilliams Identity (MacWilliams 1963) of coding theory fame. The Tutte polynomial is now of wide interest and appeal to the broader mathematical community who have found it lurking disguised in numerous areas of mathematics. Despite its present prominence, few are aware of how the Tutte polynomial provides another beautifully simple proof of a second celebrated duality theorem from coding theory, namely Weis Duality Theorem (Wei 1991). This proof, due to Duursma (2004), deserves better exposure, so this talk will present Duursmas proof.

Colloquium of Mathematics and Systems Analysis

**Prof. Kari Astala (University of Helsinki)**

**Self-similar sets (Teaching Demonstration for adjunct professor position)**

Tuesday 25 April 2017, 14:15, M1 (M232)

**Elina Numminen (University of Helsinki)**

**The epidemics of a wild plant pathogen obey the laws of graph theory**

Tuesday 25 April 2017, 10:15, M237

The ecology and evolution of any pathogen is driven by the availability and dynamics of its host. For instance the surrounding landscape, and how it creates transmission barriers and highways can critically constrain the epidemiological dynamics. In my presentation, I will present analyses on a densely sampled longitudinal genomic dataset in order to assess the features driving spatio-temporal diversity and transmission dynamics of an obligate fungal plant pathogen, Podosphaera plantaginis in the Åland archipelago, the natural patchy metapopulation it occurs in. I utilize modern statistical inference techniques and mathematical graph theory to assess how different landscape features, especially roads impact the spread of this wind-dispersing pathogen and how this results in what seems at first as cryptic patterns of pathogen diversity. Our results show a prime example of analysing epidemics with the help of network concepts, an approach which often is tempting, yet inplausible due to lack of knowledge on the transmission network.

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**Prof. Daniele Boffi (Universitá di Pavia)**

**Finite elements for elliptic problems: a priori error estimates (Teaching demonstration for adjunct professor position)**

Monday 24 April 2017, 11:00, M2 (M233)

**Casimir Lindfors PERUTTU CANCELLED**

Friday 21 April 2017, 14:15, M3 (M234)

Harmonic analysis and PDE seminar

**Axel Målqvist, Chalmers University of Technology**

**Localization of elliptic multiscale problems**

Thursday 20 April 2017, 10:00, M2 (M233)

Seminar in Applied Mathematics

**Prof. Samuli Siltanen (University of Helsinki)**

**Electrical impedance tomography imaging via the Radon transform**

Wednesday 19 April 2017, 12:15, M3 (M234)

Seminar on analysis and geometry

**Ville Tengvall (University of Jyväskylä)**

**A Sobolev homeomorphism that cannot be approximated by diffeomorphisms in $W^{1,p}$**

Wednesday 12 April 2017, 12:15, M3 (M234)

Joint work with Daniel Campbell and Stanislav Hencl

Seminar on analysis and geometry

**Liam Solus (KTH)**

**Learning Bayesian Networks via Edge Walks on DAG Associahedra**

Tuesday 11 April 2017, 10:15, M237

The focus of this talk will be an application of convex polytopes to causal inference. Graphical models based on directed acyclic graphs (DAGs), also known as Bayesian networks, are used to model complex cause-and-effect systems across a vast number of research areas including computational biology, epidemiology, sociology, and environmental management. A DAG model is family of joint probability distributions over the nodes of a DAG G that entail a set of conditional independence (CI) relations encoded by the nonedges of G. A fundamental problem in causality is to learn an unknown DAG G based only on a set of observed conditional independence relations. Since multiple DAGs can encode the same set of CI relations, a property termed Markov equivalence, the goal is to identify efficient algorithms that consistently recover a DAG within the correct Markov equivalence class. We will describe a pair of greedy algorithms for DAG model selection that operate via edge walks on a family of generalized permutohedra called DAG associahedra. We will present consistency guarantees for these algorithms, and compare them with the more classical approaches to Bayesian model selection in both efficiency and strength of satisfied identifiability assumptions. To better understand the efficiency of these new algorithms, we study two new generating functions associated to a graph which share rich connections with classic combinatorial structures.

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**Juulia Happonen**

**Theory Behind Regulatory Capital Formulae**

Monday 10 April 2017, 15:15, M2 (M233)

**Matias Vestberg**

**Up to Your Neck in Shallow Water**

Friday 31 March 2017, 14:15, M3 (M234)

We take a plunge into the world of shallow water dynamics as described by the diffusive wave approximation. This nonlinear parabolic PDE is studied using De Giorgi and Moser type iteration techniques to obtain local boundedness of weak solutions. We also present fundamental solutions in the flat case and make some remarks regarding the existence of solutions.

Harmonic analysis and PDE seminar

**Matthew Romney (University of Illinois)**

** Existence of quasiconformal parametrizations of metric surfaces with improved dilatation**

Wednesday 29 March 2017, 12:15, M3 (M234)

The measurable Riemann mapping theorem is the fundamental existence theorem for homeomorphic solutions to the Beltrami equation in the plane, i.e. quasiconformal mappings, and shows the flexibility of such mappings. More particularly, it gives hope for a positive answer to this question: given a metric space (X,d) quasiconformally homeomorphic to a simply connected domain in the complex plane, can one find a quasiconformal mapping from X onto the same domain which improves the dilatation to within a fixed universal bound? If so, what it the sharp value? This talk will discuss a solution of this problem.

Seminar on analysis and geometry

**Christian Haase (Freie Universität Berlin)**

**Finiteness Theorems for Lattice Polytopes**

Tuesday 28 March 2017, 15:15, M1 (M232)

A lattice polytope is the convex hull of finitely many points all whose coordinates are integers. These fundamental geometric objects appear in a variety of mathematical fields like combinatorics or algebraic and symplectic geometry, with applications ranging from optimization to statistics to mathematical physics in string theory.
Considerable effort has gone into several classification projects for classes of lattice polytopes. All these classifications are based on theorems, stating that the class under consideration has only finitely many elements, or at least that certain parameters are bounded in the class.
In this talk, I will give an overview of my favorite finiteness theorems, starting in the 19th century with Pick and Minkowski and ending with recent developments to be published in 2017+.

Colloquium of Mathematics and Systems Analysis

**Johan Wästlund (Chalmers University of Technology)**

**Optimization Through Games**

Tuesday 28 March 2017, 10:15, M237

Further information

I will discuss a couple of different methods for analyzing optimization problems on large random structures. In particular I will show how some results can be derived by inventing two-person games whose optimal strategies encode the solution to problems like minimum weight matching on weighted graphs.

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**Matthias Grezet**

**Rank distribution and generalized weights of Delsarte codes**

Wednesday 22 March 2017, 15:15, M3 (M234)

Delsarte codes and, in particular, all rank-metric codes play an important role in network coding to correct errors and secure an adversarial channel. The purpose of this talk is to study the properties inherent to Delsarte codes via algebraic and combinatoric methods. We will first present an analogue of the Singleton bound theorem for this context. Then, we will talk about the rank distribution of a specific code. Finally, we will introduce a new invariant, called the set of generalized weights, and see how it can characterize different type of codes

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Mikko Stenlund (University of Jyväskylä)**

**Stein's method for dynamical systems**

Wednesday 22 March 2017, 12:15, M3 (M234)

We present an adaptation of Steins method of normal approximation to the study of dynamical systems. The method yields a convergence rate essentially with the same amount of work as the central limit theorem, together with a multiplicative constant that can be computed directly from the assumptions.

Seminar on analysis and geometry

**Estibalitz Durand Cartagena (UNED, Spain)**

**∞-harmonic functions on metric measure spaces**

Friday 17 March 2017, 15:00, M3 (M234)

Harmonic analysis and PDE seminar

**Prof. Henry Segerman (Oklahoma State University)**

**3D Shadows: Casting Light on the Fourth Dimension**

Thursday 16 March 2017, 18:30, Heureka

Our brains have evolved in a three-dimensional environment, and so we are very good at visualising two- and three-dimensional objects. But what about four-dimensional objects? The best we can really do is to look at three-dimensional "shadows". Just as a shadow of a three-dimensional object squishes it into the two-dimensional plane, we can squish a four-dimensional shape into three-dimensional space, where we can then make a sculpture of it. If the four-dimensional object isn't too complicated and we choose a good way to squish it, then we can get a very good sense of what it is like. We will explore the sphere in four-dimensional space, the four-dimensional polytopes (which are the four-dimensional versions of the three-dimensional polyhedra), and various 3D printed sculptures, puzzles, and virtual reality experiences that have come from thinking about these things. I talk about these topics and much more in my new book, "Visualizing Mathematics with 3D Printing".

Mathematics and Arts Colloquium / Kirsi Peltonen

**Prof. Henry Segerman (Oklahoma State University)**

**"Design of 3D printed mathematical art" **

Wednesday 15 March 2017, 16:00, B-Hall

When visualising topological objects via 3D printing, we need a three-dimensional geometric representation of the object. There are approximately three broad strategies for doing this: "Manual" - using whatever design software is available to build the object by hand; "Parametric/Implicit" - generating the desired geometry using a parametrisation or implicit description of the object; and "Iterative" - numerically solving an optimisation problem.
The manual strategy is unlikely to produce good results unless the subject is very simple. In general, if there is a reasonably canonical geometric structure on the topological object, then we hope to be able to produce a parametrisation of it. However, in many cases this seems to be impossible and some form of iterative method is the best we can do.
I will discuss these matters with many examples, including visualisation of four-dimensional polytopes (using orthogonal versus stereographic projection) and Seifert surfaces (comparing my work with Saul Schleimer with Jack van Wijk's iterative techniques).
I will also describe some computational problems that have come up in my 3D printed work, including the design of 3D printed mobiles (joint work with Marco Mahler), "Triple gear" and a visualisation of the Klein Quartic (joint work with Saul Schleimer), and hinged surfaces with negative curvature (joint work with Geoffrey Irving).

Mathematics and Arts Colloquium / Kirsi Peltonen

**Taoufiq Damir**

**The bounded gaps between primes (NOTE unusual time!)**

Wednesday 15 March 2017, 14:15, M3 (M234)

The aim of the talk is to present a survey on the main ideas developed by Selberg, Goldston, Pintz, Yildirim, Maynard and Tao in order to prove the bounded gap between primes, namely the existence of infinitely many primes p such that p+M is prime, where M is some positive even integer.
The talk will be in two parts. The first part will be mostly dedicated to an introduction to elementary Analytic number theory and Sieve theory, then in the second part we will sketch the proofs of the main results leading to the bounded gaps.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Eero Ruosteenoja (University of Jyväskylä)**

**Tug-of-war games**

Wednesday 15 March 2017, 12:15, M3 (M234)

Ten years ago Peres, Schramm, Sheffield and Wilson introduced two-player zero-sum stochastic games that they called tug-of-war games. They showed that these games have close connections to certain nonlinear PDEs of p-Laplacian and infinity Laplacian type. I will give an overview of the topic. No background knowledge of probability theory is needed.

Seminar on analysis and geometry

**Petteri Kaski**

**Directed Hamiltonicity and Generalized Laplacians**

Tuesday 14 March 2017, 10:15, M237

The (symbolic) Laplacian matrix of a graph G is a rich source of algebraic combinatorics. Our interest is on directed graphs, where by puncturing the Laplacian at a vertex r and taking the determinant, we obtain, by the classical (directed) MatrixTree Theorem, a symbolic generating function for the spanning out-branchings of G rooted at r. In this talk we seek to understand Laplacian determinants via the algebraic combinatorics of incidence assignments that we can then turn into new randomized algorithm designs for, e.g., the Hamiltonian cycle problem on directed graphs.
We present two such algorithm designs. First, for any constant 0 < \lambda < 1 and any prime p, we present a randomized algorithm that, given an n-vertex directed graph G as input, counts the number of Hamiltonian cycles modulo p^{\lfloor(1-\lambda)n/(3p)\rfloor} in expected time c^n for a constant c<2 that depends only on p and \lambda. Second, we present a randomized algorithm that with a negligible probability of a false negative detects a directed Hamiltonian cycle in a given n-vertex directed graph G in time O*(3^{n-\alpha(G)}) and space polynomial in n, where \alpha(G) is the size of a maximum independent set in G.
This is joint work with Andreas Björklund (Lund University) and Ioannis Koutis (University of Puerto Rico).

ASci Large Structures Seminar / Freij-Hollanti, Kubjas

**Jonas Tölle**

**The ***p*-Laplace evolution equation as *p* tends to 1: Mosco convergence and convergence of solutions

Friday 10 March 2017, 14:15, M3 (M234)

Harmonic analysis and PDE seminar

**Casimir Lindfors**

**Regularity for nonlinear parabolic PDEs with general growth**

Thursday 09 March 2017, 15:40, M3 (M234)

Doctoral student talk / Engström

**Pekka Lehtelä**

**Generalized solutions to the porous medium equation**

Thursday 09 March 2017, 15:15, M3 (M234)

Doctoral student talk / Engström

**Hoa Ngo**

**Information spreading in large networks**

Thursday 09 March 2017, 14:40, M3 (M234)

Doctoral student talk / Engström

**Armando Gutiérrez**

**The horoboundary of metric spaces and its role in dynamics**

Thursday 09 March 2017, 14:15, M3 (M234)

Doctoral student talk / Engström

**Juho Andelmin**

**Solving the green vehicle routing problem and allocating resources based on efficiency analysis**

Thursday 09 March 2017, 13:40, M3 (M234)

Doctoral student talk / Engström

**Yrjänä Hynninen**

**Mathematical methods for medical decision making**

Thursday 09 March 2017, 13:15, M3 (M234)

Doctoral student talk / Engström

**Anton von Schantz**

**Modeling crowd dynamics in an evacuation as a multiagent system**

Thursday 09 March 2017, 11:40, M205

Doctoral student talk / Engström

**Heikki Puustinen**

**Network optimization and simulation in military air mission planning**

Thursday 09 March 2017, 11:15, M205

Doctoral student talk / Engström

**Pekka Laitila**

**Facilitating expert elicitation of probabilities for Bayesian networks**

Thursday 09 March 2017, 10:40, M205

Doctoral student talk / Engström

**Mikko Harju**

**Spatial decision analysis with incomplete information**

Thursday 09 March 2017, 10:15, M205

Doctoral student talk / Engström

**Tom Gustafsson**

**Numerical methods for lubrication and elastic contact**

Wednesday 08 March 2017, 16:40, M205

Doctoral student talk / Engström

**Mikael Laaksonen**

**Spectral inverse iteration for stochastic eigenvalue problems**

Wednesday 08 March 2017, 16:15, M205

Doctoral student talk / Engström

**Lauri Mustonen**

**Surrogate models and computational inverse problems**

Wednesday 08 March 2017, 15:40, M205

Doctoral student talk / Engström

**Anna-Lena Horlemann-Trautmann**

**Symbol Erasures in Random Network Coding**

Wednesday 08 March 2017, 15:15, M3 (M234)

In random network coding we want to communicate over a noisy network channel with one sender and several receivers, where all receivers want to get the same information from the sender. We encode the data before sending it through the network, so the receivers can reconstruct the original data from the corrupted data. From a mathematical point of view, classical tools used in this context are projective spaces over finite fields.
In this talk we will give an introduction to random network coding and explain how the classical model by Kötter and Kschischang treats erasures during the communication. Then we give an alternative model and show that, depending on the network topology, the classical or the alternative model can be advantageous for erasure correction.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Vesa Kaarnioja**

**On the structure and eigenvalues of lattice-theoretic tensors**

Wednesday 08 March 2017, 15:15, M205

Doctoral student talk / Engström

**Prof. Kaie Kubjas (Aalto)**

**Matrices, photos and Netflix**

Wednesday 08 March 2017, 15:00, TU1

LUMA-luennot lukiolaisille / Kangaslampi

**Juha Kuortti**

**Spectral characteristics of the vocal tract**

Wednesday 08 March 2017, 14:40, M205

Doctoral student talk / Engström

**Diana Apetrei **

**Advances in multicentric calculus**

Tuesday 07 March 2017, 16:40, M203

Doctoral student talk / Engström

**Tiina Vesanen**

**Numerical aspects of multicentric calculus**

Tuesday 07 March 2017, 16:15, M203

Doctoral student talk / Engström

**Antti Ojalammi**

**On exterior acoustics and speech modelling**

Tuesday 07 March 2017, 15:40, M203

Doctoral student talk / Engström

**Ferdinand Blomqvist**

**The closest vector problem in high dimension**

Tuesday 07 March 2017, 15:15, M203

Doctoral student talk / Engström

**Lauri Perkkiö**

**Differential material parameter in finite element analysis**

Tuesday 07 March 2017, 10:40, M205

Doctoral student talk / Engström

**Amaro Barreal**

**Lattices for physical layer network coding**

Tuesday 07 March 2017, 10:15, M205

Doctoral student talk / Engström

**Prof. Petteri Kaski**

**Arithmetic circuits for multilinear tasks (45min, NOTE unusual time!)**

Wednesday 01 March 2017, 17:15, M3 (M234)

This talk will give a brief introduction to the design of arithmetic circuits for solving computational tasks, with a focus on multilinear tasks and designs, such as fast matrix multiplication and fast polynomial multiplication in the bilinear case, and higher-order designs such as determinants, permanents, and our recent design (with Andreas Björklund) for the \binom{6}{2}-linear form, which e.g. captures the task of counting the 6-cliques in a given graph.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Clifford Gilmore (University of Helsinki)**

** Growth rates of frequently hypercyclic harmonic functions**

Wednesday 01 March 2017, 12:15, M3 (M234)

The notion of frequent hypercyclicity stems from ergodic theory and has been an active area of research since it was introduced by Bayart and Grivaux (2004).
Many natural bounded linear operators are frequently hypercyclic, for instance the differentiation operator on the space of entire functions.
We will begin by recalling some basic examples and the pertinent notions of frequent hypercyclicity.
We then consider the partial differentiation operator acting on the space of harmonic functions on R^n.
Our primary goal is to identify sharp growth rates, in terms of the L^2-norm, of harmonic functions that are frequently hypercyclic vectors for the basic partial differentiation operator. This answers a question posed by Blasco et al. (2010).
This is joint work with Eero Saksman and Hans-Olav Tylli.

Seminar on analysis and geometry

**Eero Saksman (University of Helsinki)**

**The Riemann zeta function meets Gaussian multiplicative chaos**

Tuesday 28 February 2017, 15:15, M1 (M232)

We begin by recalling some basic properties of the Riemann zeta
function and its relevance to number theory. Later on we describe it's chaotic nature on the critical strip. The final theme to be discussed is the functional statistics of the zeta function on the critical line. The last mentioned topic is based on joint work with Christian Webb (Aalto).

Colloquium of Mathematics and Systems Analysis

**Dario Gasbarra (University of Helsinki)**

**Estimating passengers travel lenghts distributions in Helsinki commuter trains**

Tuesday 28 February 2017, 10:15, M237

A local train stops at stations 0, 1, . . . , T and passengers
getting in and out a wagon are counted at every station. Since
passengers are not labeled and with the same ticket it is possible to
travel across all the line, when a passenger gets off the train, it is
not known from which station he started. The distribution of the
individual trips satisfies a constrained system of linear
diophanitine equations, which typically is underdetermined.
Nevertheless we show that, when i.i.d. realizations are
collected , the probability distribution of the travel lengths,
depending on the passengers departure station is identifiable.
We use a Bayesian data augmentation, where the the
distribution of the individual trips is the hidden variable. The
posterior distribution is explored by the fiber walk Markov chain
Monte introduced by Diaconis and Sturmfels 1998.

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**Ivo Kubjas**

**Two-Party Function Computation**

Wednesday 22 February 2017, 15:15, M3 (M234)

Assume a distributed system with two users, each user possesses a collection of binary strings. We introduce a new problem termed function computation on the reconciled data, where the users seek to compute a function on the union of their collections. Trivially, the users could just exchange their collections and then compute the value of the function. We see that any deterministic protocol can not perform much better in terms of bits communicated between the users. Namely, we show that any protocol that computes a sum and a product of reconciled sets of non-negative integers has to communicate at least 2^n + n - 1 and 2^n + n - 3 bits in the worst-case scenario, respectively. We also consider other approaches for estimating the communication complexity of the protocols. We establish connections to other problems in computer science, such as set disjointness and finding the intersection, yielding a variety of additional bounds. We present a randomized protocol, which is based on use of a family of hash functions and analyze its characteristics.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Aleksis Koski (University of Jyväskylä)**

**Nonlinear Beltrami equations: Improved regularity**

Wednesday 22 February 2017, 12:15, M3 (M234)

The Beltrami equation is a planar elliptic equation generalizing the Cauchy-Riemann equations for holomorphic functions. In this talk we will introduce this equation along with its linear and nonlinear variants. We will explain some recent results on the nonlinear Beltrami equation related to the regularity of solutions, uniqueness properties of normalized solutions and the positivity of the Jacobian. We also study connections to divergence-type equations. The talk is based on joint work with Kari Astala, Albert Clop, Daniel Faraco and Jarmo Jääskeläinen.

Seminar on analysis and geometry

**Riikka Kangaslampi**

**Graph Curvature**

Tuesday 21 February 2017, 10:15, M237

We introduce the Ollivier-Ricci curvature of graphs, which is a curvature notion defined on the edges of the graph. As an example we calculate the curvature of the 1-sceleton of the icosidodecahedron, i.e., the icosidodecahedral graph. We also define briefly Bakry-Émery curvature, which is a curvature notion defined on the vertices of a graph, and discuss connections and differences of Ollivier-Ricci curvature and the normalized Bakry-Émery curvature. We calculate the Ollivier-Ricci curvature of strongly regular graphs of girth four and five and compare it with the corresponding Bakry-Émery curvatures, and derive a partial comparison result between both curvature notions in the family of triangle free regular graphs.

Combinatorics / Engström

**Lisa Beck (Augsburg)**

**On the minimization of convex, variational integrals of linear growth**

Friday 17 February 2017, 14:15, M3 (M234)

We study the minimization of functionals

$$ \int_\Omega f(Du) \, dx $$

with a convex integrand $f$ of linear growth (such as the area
integrand), among all functions in the Sobolev space $W^{1,1}$ with
prescribed boundary values. Due to insufficient compactness properties
of these Dirichlet classes, the existence of solutions does not follow
in a standard way by the direct method in the calculus of variations and
in fact might fail, as it is well-known already for the non-parametric
minimal surface problem.
Assuming radial structure, I explain for the scalar case a necessary and
sufficient condition on the integrand such that the Dirichlet problem is
solvable, in the sense that a Lipschitz solution exists for any regular
domain and all prescribed regular boundary values, via the construction
of appropriate barrier functions in the tradition of Serrin. Since for
the general case solutions exist only in a suitably generalized sense, I
then discuss the extension of the original functional to the space of
functions of bounded variation via relaxation and give some results
concerning the regularity and uniqueness of such generalized solutions.
The results presented in this talk are based on joined projects with
Miroslav Bulíček (Prague), Erika Maringová (Prague), and Thomas Schmidt
(Hamburg).

Harmonic analysis and PDE seminar

** Kaie Kubjas **

**Algebraic methods for maximum likelihood estimation**

Thursday 16 February 2017, 12:15, M1 (M232)

Given data and a statistical model, the maximum likelihood estimate is the point of the statistical model that maximizes the probability of observing the data. In this talk, I will give examples how algebraic methods can help to solve the maximum likelihood estimation problem.

Women in Mathematics Colloquium + Free lunch/ Viveka Erlandsson

**Prof. Sihem Mesnager (University of Paris 8)**

**Linear codes from bent functions over finite fields**

Thursday 16 February 2017, 10:15, M2 (M233)

Error correcting codes are widely studied by researchers and employed by engineers. They have long been known to have applications in computer and communication systems, data storage devices (starting from the use of Reed Solomon codes in CDs) and consumer electronics. A lot of progress has been made on the constructions of linear codes with few weights. Such codes have applications in secret sharing authentication codes, association schemes and strongly regular graphs.
Certain special types of functions over finite fields are closely related to linear or nonlinear codes. In the past decade, a lot of progress on interplays between special functions and codes has been made. In particular, APN functions, planar functions, Dickson polynomials, and q-polynomials were employed to construct linear codes with optimal or almost optimal parameters. Recently, several new approaches to constructing linear codes with special types of functions were proposed, and a lot of linear codes with excellent parameters were obtained.
Bent functions are maximally nonlinear Boolean functions. They were introduced by Rothaus in the 1960's and initially studied by Dillon as early as 1974 in his Thesis. The notion of bent function has been extended in arbitrary characteristic. For their own sake as interesting combinatorial objects, but also for their relations to coding theory (e.g. Reed-Muller codes, Kerdock codes, etc.), combinatorics (e.g. difference sets), design theory, sequence theory, and applications in cryptography (design of stream ciphers and of S-boxes for block ciphers), bent functions have attracted a lot of research for the past four decades.
It is well-known that Kerdock codes are constructed from bent functions. Very recently, some authors have highlighted that bent functions lead to the construction of interesting linear codes (in particular, linear codes with few weights). This talk is devoted to linear codes from bent functions in characteristic $p$. We shall present the state of the art as well as our recent contributions in this topic. We will present two generic constructions of linear codes involving special functions and investigate constructions of good linear codes based on the generic constructions involving bent functions over finite fields. More specifically, we shall give more details on our recent results on linear codes with few weights from weakly regular bent functions based on a generic construction.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Piermarco Milione**

**Quaternion Algebras, Arithmetic, and Applications. **

Wednesday 15 February 2017, 15:15, M2 (M233)

Quaternion algebras are a wide generalization of the famous Hamilton's quaternions, and they include matrix algebras as special cases. They are the first examples of non-commutative algebras, and it is precisely the lack of commutativity which makes them appealing for many applications, not only in mathematics. In this talk we give an introduction to the arithmetic theory of quaternion algebras, stressing analogies and differences with the arithmetic of number fields. We take as archetype the order of Hurwitz's quaternions, which is the analog of the ring of Gaussian integers, in this non-commutative context. Finally, we also give a brief overview of some results, joint with L. Amorós, which are obtained in the study of the arithmetic in certain quaternion algebras and the corresponding bad reductions of certain families of Shimura curves.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Martina Aaltonen (University of Helsinki)**

**Non-manifold monodromy spaces of branched covering maps between manifolds.**

Wednesday 15 February 2017, 12:15, M3 (M234)

Abstract: Berstein and Edmonds showed that every proper branched cover F between manifolds is a factor of a branched covering orbit map. The domain of the orbit map is a locally connected and locally compact Hausdorff space and called the monodromy space of F. For proper branched covers between 2-manifolds the monodromy space is known to be a manifold. We show that this does not generalize to dimension 3 by constructing a self-map of the 3-sphere for which the monodromy space is not a locally contractible space.

Seminar on analysis and geometry

**Alex Jung**

**Compressed Sensing for Learning from Big Data over Networks **

Tuesday 14 February 2017, 10:15, M237

In this talk, I present some of our most recent work on applying tools from compressed sensing to (semi-supervised) machine learning with massive network-structured datasets, i.e., big data over networks. We expect the use of compressed sensing ideas game changing as it was for digital signal processing. In particular, I will present a sparse label propagation algorithm which efficiently learns the labels for data points based on the availability of a few labeled training data points. This algorithm is inspired by compressed sensing recovery methods and allows for a simple sufficient condition on the network structure which guarantees accurate learning.

AScI Large Structures Seminar / Freij-Hollanti, Kubjas

**Hannes Luiro (University of Jyväskylä)**

**W^{1,1}-problem for the maximal operator**

Wednesday 08 February 2017, 12:15, M3 (M234)

**Prof. Shouhei Honda (Tohoku University)**

**Geometric analysis on RCD-spaces and applications**

Wednesday 01 February 2017, 12:15, M3 (M234)

In this talk, we discuss metric measure spaces with Ricci curvature bounded from below, so called RCD-spaces.
Alexandrov spaces, weighted Riemannian manifolds, and measured Gromov-Hausdorff (mGH) limit spaces of Riemannian manifolds with Ricci curvature bounded from below give typical examples of such spaces.
The main purpose of this talk is to introduce several stability results of Sobolev/BV-functions with respect to mGH-convergence.
Applications include a new spherical suspension theorem related to the $p$-Laplacian ($1 \le p \le \infty$) on an RCD-space, which is new even for smooth manifolds.
Weyl's law on an RCD-space is also discussed if there is sufficient time.
This talk is based on joint works with L. Ambrosio (arXiv:1605.07908), and with L. Ambrosio - D. Tewodrose (arXiv:1701.03906).

Seminar on analysis and geometry

**Jarkko Kari (University of Turku)**

**An Algebraic Geometric Approach to Multidimensional Symbolic Dynamics**

Tuesday 31 January 2017, 15:15, U1 (U154)

We study low complexity multidimensional words and subshifts using tools of algebraic geometry. The low complexity assumption is that, for some finite shape D, the word or the subshift has at most |D| distinct patterns of shape D. We express words as multivariate formal power series over integers and notice that the low complexity assumption implies that there is an annihilating polynomial: a polynomial whose formal product with the power series is zero. We prove that the word must then be a sum of periodic words over integers, possibly with unbounded values. As a specific application of the method we obtain an asymptotic version of the well-known Nivat's conjecture: we show that a two-dimensional word that has low complexity with respect to arbitrarily large rectangles D must be periodic.

Colloquium of Mathematics and Systems Analysis

**Ragnar Freij-Hollanti (Aalto University)**

**Reed-Solomon codes and Private Information Retrieval**

Tuesday 31 January 2017, 10:15, M237

Private information retrieval (PIR) addresses the question of how to retrieve data items from a database without disclosing information about the identity of the data items retrieved. The area has received renents of all thewed attention, after a model of coded PIR was introduced. Here the files are distributed over the servers according to a storage code, so there is no assumption of the conte servers being identical. We present a general framework for PIR from arbitrary coded databases, that allows one to adjust the scheme according to the suspected collections of colluding servers. We will survey capacity results and constructions in this area.
In the second half of the talk, we introduce generalized Reed-Solomon (GRS) codes, and discuss some of their algebraic properties that are desirable in data storage applications. We show that our PIR schemes have optimal performance if the storage code is a generalized Reed-Solomon code. We compare our constructions to known capacity bounds for coded PIR, and introduce a new model of coded PIR for caching, in which we conjecture that our constructions are optimal. This is joint work with Salim El Rouayheb, Oliver Gnilke, Camilla Hollanti, David Karpuk, and Razane Tajjedine.

Large Structures Seminar / Freij-Hollanti, Kubjas

**Riikka Kangaslampi (Aalto)**

**Fraktaalien matematiikkaa**

Wednesday 25 January 2017, 15:00, TU1

LUMA-luennot lukiolaisille / Kangaslampi

**Prof Juan Souto (Universite de Rennes 1)**

**Critical sets of smooth functions on R^3**

Wednesday 25 January 2017, 10:15, M3 (M234)

Note the exceptional time !
Abstract: While it is classical, and otherwise easy to see, that each closed set of R^n is the set of zeroes of a smooth function, much less is known about which sets arise as sets of critical points. In this talk I will discuss some results in 3 dimensions.

**Rami Luisto (University of Jyväskylä)**

**Stoilow's theorem revisited**

Wednesday 18 January 2017, 12:15, M3 (M234)

Stoı̈lows theorem from 1928 states that a continuous, light,
and open mapping between surfaces is a discrete map with a discrete branch set. This result implies that such mappings between orientable surfaces are locally modelled by complex power mappings and admit a holomorphic factorization. The purpose of the talk is to give a proof of this classical
theorem having the readers interested in discrete and open mappings in mind.

Seminar on analysis and geometry

**Peter Lindqvist (Trondheim)**

**Non-linear Eigenvalue Problems**

Friday 13 January 2017, 14:15, M3 (M234)

Harmonic analysis and PDE seminar

**Prof. Xiao Zhong (University of Helsinki ) **

**On the Euler-Lagrange equation of a functional by Pólya and Szegö**

Wednesday 11 January 2017, 12:15, M3 (M234)

I will talk about a conjecture of Pólya and Szegö on minimal electrostatic capacity sets in convex shape optimization.
The functional, associated to the conjecture, involves capacity and perimeter. We will focus on the regularity properties of generalized solutions of the corresponding Euler-Lagrange equation. This is joint work with Nicola Fusco.

Seminar on analysis and geometry

**Kristian Ranestad (Oslo University)**

**Symmetric and Lagrangian degeneracy loci: Kummer surfaces and hyperkähler fourfolds.**

Tuesday 10 January 2017, 11:15, M240

Kummer quartic surfaces may be constructed as both symmetric and Lagrangian degeneracy loci. I shall discuss these constructions, relations between them, and similar constructions of hyperkähler fourfolds, in a report on common work with Atanas Iliev, Grzegorz and Michal Kapustka.

**Prof. Olavi Nevanlinna**

**On computing spectra of bounded operators in separable Hilbert spaces**

Monday 09 January 2017, 14:15, M2 (M233)

Seminar on Applied Mathematics

**David Karpuk**

** Lattices for Reliable and Secure Wireless Communication**

Wednesday 04 January 2017, 15:15, Y228a

Wireless communications is ubiquitous in modern life, and the mathematics of communications engineering is fundamental to everything from large cellular networks to small Wi-Fi networks. In wireless communications, data is often represented as a finite subset of R^n called a constellation, and choosing a constellation cleverly can increase reliability of transmission or protect against malicious eavesdroppers. Constellations carved from lattices, that is, discrete subsets of R^n, are especially useful as the underlying algebraic structure of lattices allows for analysis of certain error probabilities. In this talk we will begin with a survey of practical lattice constructions, discuss our recent work concerning lattices from number-theoretic constructions and Hadamard matrices, and discuss future work regarding applications of so-called well-rounded lattices.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Juhan Aru (ETH Zürich)**

**Balancing random sums of vectors**

Tuesday 20 December 2016, 15:15, M1 (M232)

Further information

Stochastic Sauna 2016

**Lasse Leskelä (Aalto)**

**Clustering coefficients in large directed graphs**

Tuesday 20 December 2016, 15:15, M1 (M232)

Further information

Stochastic Sauna 2016

**Alex Jung (Aalto)**

**Graphical Model Selection for Learning from Big Data over Networks**

Tuesday 20 December 2016, 14:00, M1 (M232)

Further information

Stochastic Sauna 2016

**Jani Lukkarinen (Univ. Helsinki)**

**Summability of joint cumulants of nonindependent lattice fields**

Tuesday 20 December 2016, 13:00, M1 (M232)

Further information

Stochastic Sauna 2016

**Sara Taskinen (Univ. Jyväskylä)**

**Generalized linear latent variable models in the analysis of multivariate abundance data**

Tuesday 20 December 2016, 11:15, M1 (M232)

Further information

Stochastic Sauna 2016

**Tommi Sottinen (Univ. Vaasa)**

**Prediction of Gaussian processes**

Tuesday 20 December 2016, 10:15, M1 (M232)

Further information

Stochastic Sauna 2016

**Christophe Prange (Université de Bordeaux)**

**Quantitative analysis of boundary layers in periodic homogenization**

Wednesday 14 December 2016, 10:00, M2 (M233)

In this talk I will focus on new quantitative estimates concerning the homogenization of elliptic systems with oscillating Dirichlet boundary conditions. The results are joint with Scott Armstrong, Tuomo Kuusi and Jean-Christophe Mourrat.

Seminar on analysis and geometry / Tuomo Kuusi

**Bernd Schwarzenbacher (TU Wien)**

**Algebraic Multigrid for Maxwell Equations**

Monday 12 December 2016, 14:15, M2 (M233)

Seminar in Applied Mathematics

**Rico Zacher (Ulm University)**

**Li-Yau and Harnack inequalities on graphs**

Friday 09 December 2016, 14:30, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Bartłomiej Dyda (Politechnika Wrocławska)**

**On function spaces and extension results for nonlocal Dirichlet problems**

Friday 09 December 2016, 13:30, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Tadele Mengesha (University of Tennessee)**

**Weighted Sobolev regularity of weak solutions of degenerate/singular elliptic problems**

Friday 09 December 2016, 10:45, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Cyril Imbert (CNRS)**

**Weak Harnack inequality for the Boltzmann equation without cut-off**

Friday 09 December 2016, 09:45, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Martina Hofmanová (Technische Universität Berlin)**

**Rough Gronwall Lemma and weak solutions to RPDEs**

Thursday 08 December 2016, 16:00, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Aleksandra Zimmermann (Universität Duisburg-Essen)**

**Martingale solutions for a pseudomonotone evolution equation with multiplicative noise**

Thursday 08 December 2016, 14:30, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Lisa Beck (University of Augsburg)**

**Regularization by noise for the stochastic transport equation**

Thursday 08 December 2016, 13:30, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Prof. Katie Steckles (Manchester)**

**The Hidden Maths of Technology**

Thursday 08 December 2016, 13:00, Nokia-sali (Töölö)

There are numbers all around us that make our modern lives possible. From rescuing your lost words in text messages, to taking selfies and buying products online. Katie will introduce some of the mathematical ideas used in storing, sending and manipulating information, and talk about her recent MegaPixel project - to build the world's largest digital photograph, coloured in by hand.

LUMA lecture/ Kirsi Peltonen

**Joaquim Serra (ETH Zurich)**

** Stable nonlocal phase transitions and minimal surfaces**

Thursday 08 December 2016, 10:45, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Pierre Cardaliaguet (Université Paris - Dauphine)**

**On the master equation in mean field game theory**

Thursday 08 December 2016, 09:45, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Prof. Katie Steckles (Manchester)**

**Fractals and Dynamical Systems**

Thursday 08 December 2016, 09:00, Y202a Aalto

Fractals are infinitely detailed mathematical objects, and studying them uses a range of maths, from algorithms and geometric constructions to infinite limits and different ways of measuring dimensions. Katie will introduce some mind-blowing new concepts while showing the mathematics behind natural structures and how its used in modelling.

LUMA lecture/ Kirsi Peltonen

**Jarmo Malinen**

**On resonances - the practical aspect of eigenvalue problems**

Wednesday 07 December 2016, 16:15, M2 (M233)

Seminar on Applied Mathematics

**Michael Hinz (Bielefeld University)**

**First order derivations for Dirichlet forms and applications**

Wednesday 07 December 2016, 16:00, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Eero Ruosteenoja (University of Jyväskylä)**

**Regularity for the normalized p-Poisson problem**

Wednesday 07 December 2016, 14:30, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Armin Schikorra (Albert-Ludwigs-Universität Freiburg)**

**O'Hara's knot energies and W^{1/p,p}-harmonic maps into spheres**

Wednesday 07 December 2016, 13:30, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Stefan Geiss (University of Jyväskylä)**

**On continuous time tug-of-war with noise and p-harmonic functions**

Wednesday 07 December 2016, 10:45, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Nizar Touzi (École Polytechnique)**

**Branching diffusion representation for nonlinear PDEs**

Wednesday 07 December 2016, 09:45, M1 (M232)

Further information

Workshop on Nonlinear, nonlocal problems and stochastic methods

**Alex Karrila (Aalto)**

**Boundary branches in planar uniform spanning trees**

Monday 05 December 2016, 15:15, M1 (M232)

Further information

Many planar lattice models of statistical physics are expected to have a continuum limit which is invariant under conformal maps. Physicists predict the behaviour of such models by conformal field theories (CFTs) based on this expectation. Mathematical statistical physics aims at proving rigorously conformal invariance properties.
One such lattice model is the planar uniform spanning tree (UST): given a graph which is a lattice approximation of some planar domain, one picks a tree subgraph that covers all the vertices. In a spanning tree, any two vertices are connected by a unique branch, and the scaling limit of a single branch has been identified as the SLE(2) random curve. We study the probability of observing multiple disjoint branches between given boundary vertices in a UST, as well as boundary visits of a single such boundary branch. We find explicitly these probabilities and their scaling limits, and show that the scaling limits are indeed conformally invariant and satisfy PDEs and asymptotics predicted by CFT. This is among the first verifications of third-order PDEs in the scaling limits of lattice models. In subsequent work, we believe that this will help us to identify the scaling limit of multiple boundary branches as a multiple SLE(2).

Aalto Stochastics & Statistics Seminar (Ilmonen-Kytölä-Leskelä)

**Henri Martikainen (University of Helsinki)**

**Adapted Cotlar type inequalities**

Friday 02 December 2016, 14:15, M3 (M234)

Harmonic analysis and PDE seminar

**Matias Vestberg**

**Deep Problems in Shallow Water**

Wednesday 30 November 2016, 12:15, M3 (M234)

We acquaint ourselves with a nonlinear parabolic PDE known as the diffusive wave approximation of the shallow water equations. We show local boundedness of weak solutions, thus extending previously known regularity results for the special case of a flat topography. We also obtain fundamental solutions in the flat case.

Seminar on analysis and geometry

**Prof. Leena Suhl (University of Paderborn)**

**Optimization systems to support planning processes in traffic and transportation**

Tuesday 29 November 2016, 15:15, U1 (U154)

Planning processes in large traffic and transportation systems involve complex networks with many interdependencies of components. In timetable based public transportation systems, the planning steps such as line planning, timetabling, fleet assignment, vehicle routing, crew scheduling and rostering, are usually carried out sequentially. The talk will give an overview of optimization models and solution methods to support such planning processes. Optimal vehicle rotations can, under certain conditions, be computed with low computational effort, but complex vehicle routing and crew scheduling problems cannot be solved to optimality with software available today. Thus solution methods need to involve heuristics and metaheuristics additionally to exact optimization. There is often a gap between requirements from enterprises in practice and scientific methods available in literature, and new research problems can be identified that help to close this gap. Requirements for robust schedules or integrated planning provide further challenging research questions today.

Departmental Colloquim

**Pekka Lehtelä (Aalto)**

**Unbounded supersolutions of the porous medium equation**

Friday 25 November 2016, 14:15, Kumpula C123

Harmonic analysis and PDE seminar

**David Bate (University of Helsinki)**

**Differentiability and Poincare-type inequalities in metric measure spaces.**

Wednesday 23 November 2016, 12:15, M3 (M234)

Metric measure spaces that satisfy a Poincare inequality (in the sense of Heinonen and Koskela) are a standard starting point for generalising many areas of analysis to the abstract metric setting. In particular, Cheeger proved that they satisfy a generalisation of Rademacher's differentiation theorem, which was extended by Cheeger-Kleiner to all Lipschitz functions taking values in a Banach space with the Radon-Nikodym property.
In this talk we will demonstrate the necessity of a Poincare type inequality for those metric measure spaces that satisfy Cheeger's differentiation theorem for all Lipschitz functions taking values in an RNP-Banach space. This is done by showing the existence of a rich structure of curve fragments that connect near by points, similar in nature to Semmes's pencil of curves for the standard Poincare inequality. Using techniques similar to Cheeger-Kleiner, these conditions are also sufficient. This is joint work with Sean Li.

Seminar on analysis and geometry

**Antti Ojalammi**

**Exterior domain eigenvalue problems in acoustics**

Monday 21 November 2016, 14:15, M2 (M233)

Seminar on Applied Mathematics

**Dissertation**

Janne Korvenpää

**Nonlocal nonlinear potential theory and fractional integral operators**

Friday 18 November 2016, 12:00, M1 (M232)

**Alessandro Neri (University of Zurich)**

**On the Genericity of Maximum Rank Distance and Gabidulin Codes**

Wednesday 16 November 2016, 15:15, Y229a

Codes in the rank-metric have been studied for the last four decades. For linear codes a Singleton-type bound can be derived for these codes. In analogy to MDS codes in the Hamming metric, we call rank-metric codes that achieve the Singleton-type bound MRD (maximum rank distance) codes. Since the works of Delsarte and Gabidulin we know that linear MRD codes exist for any set of parameters. The codes they describe are called Gabidulin codes.
The question, if there are other general constructions of MRD codes that are not equivalent to Gabidulin codes, has been of large interest recently. Some constructions of non-Gabidulin MRD codes are known, but many of the derived codes are not linear over the underlying field but only linear over some subfield of it.
In general, it remains an open question for which parameters non-Gabidulin MRD codes exist, and if so, how many such codes there are. We show that the properties of being MRD and non-Gabidulin are generic. This implies that over a large field extension degree a randomly chosen generator matrix generates an MRD and a non-Gabidulin code with high probability. Moreover, we give an upper bound on the respective probabilities in dependence on the extension degree.
Joint work with Anna-Lena Horlemann-Trautmann, Tovohery Randrianarisoa and Joachim Rosenthal.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Daniel Meyer (University of Jyväskylä)**

** Expanding Thurston maps and Quasispheres**

Wednesday 16 November 2016, 12:15, M3 (M234)

Abstract: A quasisphere is a metric sphere that is quasisymmetrically equivalent to the standard $2$-sphere. An
important open question is to give a characterization of quasispheres. This is closely related to Cannon's
conjecture. This conjecture may be formulated as stipulating that a group that ``behaves topologically'' as a Kleinian group ``is geometrically'' such a group. Equivalently, it stipulates that the ``boundary at infinity'' of such groups is a quasisphere.
A Thurston map is a map that behaves ``topologically'' as a rational map, i.e., a branched covering of the $2$-sphere that is postcritically finite. A question that is analog to Cannon's conjecture is whether a Thurston map ``is'' a rational map. This is answered by Thurston's classification of rational maps.
For Thurston maps that are expanding in a suitable sense, we may define ``visual metrics''. The map then is (topologically conjugate) to a rational map if and only if the sphere equipped with such a metric is a quasisphere. This is based on joint work with Mario Bonk.

Seminar on analysis and geometry

**Emanuele Ventura**

**Finite phylogenetics complexity and combinatorics of tables**

Tuesday 15 November 2016, 10:00, M3 (M234)

In this talk we show that the phylogenetic complexity, an invariant introduced by Sturmfels and Sullivant in the context of computational phylogenetics, of any finite abelian group is finite.

Combinatorics seminar

**Petri Tuisku (University of Helsinki)**

**The energy density in toric Ising model**

Monday 14 November 2016, 15:15, M1 (M232)

Further information

I calculate rigorously the scaling limit of energy density of the critical Ising model on the square lattice on the two dimensional torus. An exact formula is found for the energy density scaling limit asymptotics and thus calculations that have appeared in physics literature are validated. We also demonstrate how the energy density of the discrete Ising model is related to determinants of certain discrete Laplacians, and in the process of our calculation, we also find the scaling limit of the Ising model partition function.

Aalto Stochastics & Statistics Seminar (Ilmonen-Kytölä-Leskelä)

**Casimir Lindfors (Aalto)**

**On the Cauchy-Dirichlet problem for a general class of parabolic equations**

Friday 11 November 2016, 14:15, Kumpula C123

Harmonic analysis and PDE seminar

**Dissertation**

Helle Majander, Prof. Kim Knudsen (DTU)

**Efficient reconstruction algorithms for three-dimensional tomographic imaging**

Friday 11 November 2016, 12:00, lecture hall E

Further information

**Dr. Marzieh Arabi Kakavand (Isfahan University of Technology)**

**Simple-extending modules and semisimple-extending modules**

Wednesday 09 November 2016, 15:15, Y229a

Let M be an R-module and N be any submodule of M. A submodule C of M is called a complement of N if C is maximal in the collection of submodules Q of M such that Q\cap N=0. A submodule C of M is called complement (in M) if there is a submodule N such that C is complement of N in M.
A module M is called extending (or CS) module if every complement submodule of M is a direct summand of M. These modules were introduced by Harada and Muller in 1980 and later were studied quite extensively by many authors.
A module M is called simple-extending (semisimple-extending) if every complement of any simple (semisimple) submodule of M is a direct summand. The aim of the talk is to present some recent results concerning simple-extending and semisimple-extending modules. Namely, we will describe rings R such that every R-module M is simple-extending (semisimple-extending).

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Professor Enrico Zio (Politecnico di Milano)**

**A Modern Vision of Risk Assessment for Modern Industry**

Wednesday 09 November 2016, 14:00, M1 (M232)

Further information

As the digital, physical and human worlds continue to integrate, we experience a deep transformation in industry, which far-reaches into our lives. The 4th industrial revolution, the internet of things and big data, the industrial internet, are changing the way we design, manufacture, provide products and services. This is creating a complex network of things and people that are seamlessly connected and communicating. It is providing opportunities to make productions systems more efficient and faster, and more flexible and resilient the complex supply chains and distribution networks that tie the global economy.

In this fast-pace changing environment, the key performance indicators related to safety, reliability, maintainability of components and systems continue to play a fundamental role for industry. The innovations that are being developed have high potential of increased wellbeing and benefits, rendering everything better and smarter, but also generating new and unknown failure mechanisms, new and unknown functional and structural dependencies, and eventually new and unknown hazards and risks. On the other hand, the advancements in knowledge, methods and techniques, the increase in information sharing and data availability, offer new opportunities of analysis for optimal design, operation and maintenance of modern system engineering and industry. Then, a new revolution is in the making for addressing the challenges brought about by the new and evolved, complex systems (and systems of systems), and the innovations therein; this calls for and, at the same time, drives the advancements of new methods and tools of complex system analysis, and the extension of their applications, based on the increased knowledge, information and data (KID) available, which can improve our system behavior understanding capability in support to decision making.
In this talk, we will consider the above context and address some challenges and opportunities for complex system engineering, focusing on desired attributes of safety, reliability, maintainability but also resilience and flexibility.

Aalto Systems Forum

**David Brander (Technical University of Denmark)**

**Harmonic maps and geometric Cauchy problems**

Wednesday 09 November 2016, 12:15, M3 (M234)

t has been known since the 1990's that harmonic maps from Riemannian or Lorentzian surfaces into Symmetric spaces admit loop group generalizations of the classical Weierstrass representation (Riemannian) or d'Alembert solution of the wave equation (Lorentzian). These allow one to construct solutions to the various geometric problems that are associated to harmonic maps. The utility of these representations is obstructed by the loss of geometric information in the loop group decomposition that relates the harmonic map to the "Weierstrass" data. Recently, special types of Weierstrass data have been introduced that contain full geometric information along a curve. I will discuss applications of this technique to the construction of integrable surfaces.

Seminar on analysis and geometry

**SITE VISIT TALKS: Juliane Müller (Lawrence Berkeley National Laboratory)**

**Surrogate models in computationally expensive black-box simulation optimization**

Monday 07 November 2016, 16:15, M1 (M232)

I will discuss surrogate model algorithms for solving computationally expensive black-box optimization problems. These problems arise in many application areas, for example, in the development of simulation models for global climate studies, gas-phase chemical combustion, cosmology, and material science. The simulation models usually have several parameters that must be adjusted optimally in order to obtain a best fit or an optimal design. However, a single evaluation of the simulation model (and thus of the objective function) may take from several minutes to hours or even days, hence drastically restricting the number of parameter combinations we can try during the optimization. Derivative information is generally not available and derivative-free methods must be used. In this talk, I will focus in particular on multi-objective simulation optimization problems as well as on problems with hidden constraints.

MS site visit talks /Systems Analysis Laboratory Seminar

**Vesa Kaarnioja**

**Eigenvalue problems for high order meet and join tensors**

Monday 07 November 2016, 14:15, M2 (M233)

Seminar on Applied Mathematics

**Emil Vuorinen (University of Helsinki)**

**Characterizations of two-weight inequalities in L^p**

Friday 04 November 2016, 14:15, M3 (M234)

Harmonic analysis and PDE seminar

**Dissertation**

Jussi Kangaspunta

**Dissertation: Resource Allocation Methods for Defence and Infrastructure Systems**

Friday 04 November 2016, 12:00, H304

Further information

**SITE VISIT TALKS: Kazim Buyukboduk (Koc University, Istanbul)**

**Reciprocity Laws and p-adic analytic construction of rational points (NOTE: starts 10 sharp!)**

Friday 04 November 2016, 10:00, M2 (M233)

Our goal in this talk is to present a basic overview of reciprocity laws, starting off with that is due to Gauss and reaching all the way up to those predicted as part of Langlands' Program and Wiles' celebrated Modularity Theorem in this vein. The latter result is one of the key inputs in our recent p-adic analytic construction of a rational point of infinite order on an elliptic curve of rank one that has good supersingular reduction at p.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**SITE VISIT TALKS: Germain van Bever (Université Libre de Bruxelles)**

**Depth-based methods in statistics**

Thursday 03 November 2016, 16:15, M203

Statistical (location) depth has become more and more popular over the last decades and illustrates the growing importance geometrical methods are taking in
mathematical statistics. This talks aims at showcasing several new depth-based notions in various statistical problems. After a review of the concept of statistical depth, including the celebrated halfspace depth of Tukey (1975) and simplicial depth of Liu (1990), the following three topics will be investigated.
Despite the fact that depth indeed produces regions that provide much information on the underlying distribution, all location depths entering the general scheme of Zuo and Sering (2000) lead to convex depth regions, hence cannot deal with multimodality. A new concept of local depth is proposed (Paindaveine & VB, 2013), allowing for the analysis of, among others, multimodal distributions. Properties of this local depth concept are explored and its use is illustrated in various applications, including discriminant analysis, density estimation, etc. Robustness properties of the new concept of depth, inherited from the robust behaviour of the classical definition, are studied.
The second focus is one of the most natural applications of depth, namely classification. After a short review of the existing methods, a class of depth-based classification procedures that are of a nearest neighbour nature is proposed (Paindaveine & VB, 2015).
The corresponding nearest neighbours are identified through a symmetrised depth construction. Our classifiers enjoy (i) the robustness and affine-invariance of depth-based procedures and (ii) the good asymptotic properties of nearest-neighbour classifiers. In particular, the proposed depth-based classifiers are near-universally consistent.
Finally, recent advances (Helander et al.) in depth for data of a functional nature will be detailed. This new concept uses statistics of interest of a Pareto-like optimization approach to define a concept of depth for functional data that includes features other than location in its construction. This therefore allows to discriminate between observations that do not differ necessarily in location but also in shape, range, etc. Illustrations will be provided throughout.

MS site visit talks

**Prof. Vicki Bier (Wisconsin-Madison)**

**Game-Theoretic and Reliability Methods for Counter-terrorism and Security**

Thursday 03 November 2016, 14:00, G-111, Chydenia (Runeberginkatu 2224, Helsinki)

Further information

The routine application of reliability and risk analysis by itself is not adequate in the security domain. Protecting against intentional attacks is fundamentally different from protecting against accidents or acts of nature. In particular, an intelligent and adaptable adversary may adopt a different offensive strategy to circumvent or disable protective security measures. Game theory provides a way of taking this into account. Thus, security and counter-terrorism can benefit from a combination of risk analysis and game theory. I will discuss the use of risk and reliability analysis and game theory for defending complex systems against attacks by knowledgeable and adaptable adversaries. The results of such work yield insights into the nature of optimal defensive investments in networked systems to obtain the best trade-off between the cost of the investments and the security of the resulting systems.

Aalto Systems Forum

**Oliver Gnilke **

**Semigroup action problems in cryptography**

Wednesday 02 November 2016, 15:15, M2 (M233)

With quantum algorithms available to solve the classical discrete logarithm problem a need for different one-way functions arises. We study several alternatives to the DLP based on different semigroups and evaluate their security. A general framework for reducing semigroup action problems to smaller instances, in line with the Pohlig-Hellman attack on classical DLPs, is given.

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**SITE VISIT TALKS: Jevgenijs Ivanovs (Aarhus University)**

**Zooming in on a Lévy process at its supremum**

Wednesday 02 November 2016, 13:15, M237

Further information

Let M and τ be the supremum and its time of a Lévy process X on some finite time interval. It is shown that zooming in on X at its supremum, that is, considering (a η (X τ +t/η − M )) t∈R as η, a η → ∞, results in (ξ t ) t∈R constructed from two independent processes corresponding to some self-similar Lévy process S conditioned to stay positive and negative. This holds when X is in the domain of attraction of S under the zooming-in procedure as opposed to the classical zooming-out of Lamperti [1962]. As an application of this result we provide a limit theorem for the discretization errors in simulation of supremum and its time, which extends the result of Asmussen, Glynn, and Pitman [1995] for the Brownian motion.
In this talk I will aim at a general mathematical audience providing various illustrations of basic concepts.

MS site visit talks

**Antti Vähäkangas (University of Jyväskylä)**

**Self-improving properties of Poincaré inequalities revisited**

Wednesday 02 November 2016, 12:15, M3 (M234)

We review some of the self-improving properties of Poincaré
inequalities on doubling metric measure spaces. Our emphasis is on
novel aspects of the Keith-Zhong theorem (Annals of Mathematics, 2008).
The talk is based on a joint work with:
J. Kinnunen (Aalto University),
J. Lehrbäck (University of Jyväskylä) and
X. Zhong (University of Helsinki).

Seminar on analysis and geometry

**SITE VISIT TALKS: Matti Vihola (University of Jyväskylä)**

**Unbiased estimators and multilevel Monte Carlo**

Tuesday 01 November 2016, 14:15, M203

Further information

Multilevel Monte Carlo (MLMC) and recently proposed debiasing schemes are closely related methods for scenarios where exact simulation methods are difficult to implement, but biased estimators are available. An important application is inference with stochastic differential equation models, where the continuous-time process is difficult to simulate but time-discretized approximations are easy to implement.
A new general class of unbiased estimators is introduced, which admits earlier debiasing schemes as special cases, and accomodates new lower variance estimators based on stratification. The stratified schemes behave asymptotically like MLMC, both in terms of variance and cost, under general conditions --- essentially those that guarantee canonical square root rate of MLMC. This suggests that MLMC bias can often be eliminated entirely with small extra cost. The new schemes also admit optimisation criteria which are easy to implement in practice.
(The talk is based on arXiv:1512.01022)

MS site visit talks

**SITE VISIT TALKS: Kaie Kubjas**

**How to flatten a soccer ball**

Tuesday 01 November 2016, 10:15, M3 (M234)

In this talk, I will explain how to compute the image of a simple semialgebraic set of 3-space (soccer ball) under a polynomial map into the plane. In general cases, the boundary of the image is given by two highly singular curves. I will discuss the number of connected components of the complement of the image, maps onto convex polygons as well as connections to convex optimization. This talk is based on joint work with Pablo A. Parrilo and Bernd Sturmfels.

MS site visit talks

**Hao Wu (University of Geneva)**

**Arm Exponents for Ising and FK-Ising Model**

Monday 31 October 2016, 15:15, M3 (M234)

Further information

The introduction of SLE by Oded Schramm provides mathematicians with a new tool to study critical lattice models.
In the first part of this talk, I will discuss critical planar percolation and explain how people use SLE to derive the arm
exponents of percolation. In the second part, I will introduce SLE and general formulae on arm exponents of SLE, and
show that these formulae give us various results on the arm exponents of critical planar Ising and FK-Ising models.
Finally, I will explain some related results and some open questions.

Aalto Stochastics & Statistics Seminar (Ilmonen-Kytölä-Leskelä)

**Prof. Daniele Boffi (University of Pavia)**

**Finite element approximation of eigenvalue problems in mixed form**

Monday 31 October 2016, 14:15, M2 (M233)

Seminar on Applied Mathematics

**Patrik Norén**

**Slicing and dicing polytopes**

Friday 28 October 2016, 13:15, M2 (M233)

Using tropical convexity Dochtermann, Fink, and Sanyal proved that regular fine mixed subdivisions of Minkowski sums of simplices support minimal cellular resolutions. They asked if the regularity condition can be removed. We give an affirmative answer by a different method. A new easily checked sufficient condition for a subdivided polytope to support a cellular resolution is proved. The main tool used is discrete Morse theory.

Engström

**Petteri Kaski (Aalto University, Department of Computer Science)**

**How proofs are prepared at Camelot**

Wednesday 26 October 2016, 15:15, M2 (M233)

We study a design framework for robust, independently verifiable,
and workload-balanced distributed algorithms working on a common
input. An algorithm based on the framework is essentially
a distributed encoding procedure for a Reed--Solomon code
(each compute node is tasked to produce one or more symbols of
the codeword), which enables (a) robustness against byzantine
adversarial failures with intrinsic error-correction and
identification of failed nodes, and (b) independent randomized
verification to check the entire computation for correctness,
which takes essentially no more resources than each node individually
contributes to execute the computation. The framework also enables
smooth tradeoffs between per-node compute time and the number of
nodes used for computation.
The framework builds on recent MerlinArthur proofs of batch
evaluation of Williams [Proc. 31st IEEE Conference on Computational
Complexity (CCC'16, May 29--June 1, 2016, Tokyo), 2:117]
with the basic observation that Merlin's magic is not needed for batch
evaluation---mere Knights can prepare the independently verifiable
proof, in parallel, and with intrinsic error-correction.
[This is joint work with Andreas Björklund (Lund University).]
Link: http://dx.doi.org/10.1145/2933057.2933101

ANTA Seminar: Algebra, Number Theory, and Applications to Communications and Computing / Karpuk, Kaski, Hollanti, Greferath

**Teemu Saksala (University of Helsinki)**

**How to find the Riemannian structure from the solution mapping of a wave equation on a manifold?**

Wednesday 26 October 2016, 12:15, M3 (M234)

Further information

Seminar on analysis and geometry

**Marc Lelarge (École Normale Supérieure & INRIA)**

**Community Detection with the Non-Backtracking Operator**

Tuesday 25 October 2016, 15:15, M1 (M232)

Community detection consists in identification of groups of similar items within a population. In the context of online social networks, it is a useful primitive for recommending either contacts or news items to users. We will consider a particular generative probabilistic model for the observations, namely the so-called stochastic block model and prove that the non-backtracking operator provides a significant improvement when used for spectral clustering.

Departmental Colloquim

**Florian Kohl (FU Berlin)**

**HASE --- a program that determines the holes of (pointed) affine semigroups**

Tuesday 25 October 2016, 10:15, M3 (M234)

The question whether there exists an integral solution to the system of linear equations with non-negative constraints, Ax=b, x=> 0, where A and b have integer entries, finds its applications in many areas, such as operation research, number theory, and statistics. In order to solve this problem we have to understand the semigroup generated by the columns of the matrix A and the structure of the
``holes'' which are the difference between the semigroup generated by the
columns of the matrix A and its saturation. In this talk, we will describe the algorithm to compute the set of holes of the semigroup developed by Hemmecke, Takemura, and Yoshida and we will briefly discuss the implementation. Furthermore, we will talk about applications of our software to problems in combinatorics and statistics such as the normality property of polytopes and the set of holes for the common diagonal affect models.
This is joint work with Yanxi Li (University of Kentucky), Ruriko Yoshida
(Naval Postgraduate School), and Johannes Rauh (Leibniz Universität
Hannover)

Engström

**Hide past events**

**
**
Page content by: webmaster-math [at] list [dot] aalto [dot] fi