Norwegian version of this page

Events - Page 7

Time and place: , Room 1259 "Abels utsikt", 12th floor, Niels Henrik Abels hus, University of Oslo

A two-days meeting of the Steering Council for Centre International de Mathématiques Pures et Appliquées (CIMPA).

Time and place: , NHA107

C*-algebra seminar by Ole Brevig (University of Oslo)

Time and place: , Niels Henrik Abels hus, 9th floor

Why is deep learning so successful in many applications of modern AI? This question has puzzled the AI community for more than a decade, and many attribute the success of deep learning to the implicit regularization imposed by the Neural Network (NN) architectures and the gradient descent algorithm. In this talk we will investigate the implicit regularization of so-called linear NNs in the simplified setting of linear regression. Furthermore, we will show how this theory meets fundamental computational boundaries imposed by the phenomenon of generalized hardness of approximation. That is, the phenomenon where certain optimal NNs can be proven to exist, but any algorithm will fail to compute these NNs to an accuracy below a certain approximation threshold. Thus, paradoxically, there will exist deep learning methods that are provably optimal, but that can only be computed to a certain accuracy.

Vegard Antun is a postdoctoral fellow at the University of Oslo, department of Mathematics.

Time and place: , Abels Utsikt, 12th floor NHA

Boris Odinot, Head of Growth of grasple.com, a non-profit Ed-tech company in the Netherlands, will present about how their platform can be used to help students practice mathematic and statistics, and for teachers to monitor their students' performance.

Time and place: , Abels Utsikt (NHA 1259)
Time and place: , NHA107

QOMBINE seminar by Snorre Bergan (UiO)

Time and place: , Abels Utsikt, 12 etg. Niels Henrik Abels hus

Doctoral candidate Anton Yurchenko-Tytarenko at the Department of Mathematics will be defending the thesis Stochastic Volterra volatility models for the degree of Philosophiae Doctor.

Time and place: , Room 1119, Niels Henrik Abels hus

The Section 4 seminar for the Autumn of 2022 will be held on Thursdays from 10:15–12:00 (see the schedule)

Time and place: , NHA 1020 and Online
Time and place: , Abels Utsikt (NHA 1259)
Time and place: , Erling Sverdrups plass, Niels Henrik Abels hus, 8th floor
Time and place: , NHA B1120

Consider the singularity C^4/(Z/2), where Z/2 acts as the matrix diag(-1,-1,-1,-1). This singularity is special, in that it does not admit a crepant resolution. However, it does admit a so-called noncommutative crepant resolution, given by a Calabi-Yau 4 quiver. The moduli space of representations of this quiver turns out to share a lot of similarities with moduli spaces of sheaves over Calabi-Yau fourfolds, and it turns out that we can reuse techniques from studying moduli of sheaves to define and compute invariants of this moduli space of representations. In this talk, I will explain how these invariants can be defined, and give conjectures about the forms of these invariants. This talk is based on joint work with Raf Bocklandt.

Time and place: , Georg Sverdrups hus, Lecture hall 1

The Thoralf Skolem Memorial Lecture 2022

Time:

This mini-workshop provides young analysts in Norway with an arena to present their research and interact with their peers. 

Time and place: , Erling Sverdrups plass, Niels Henrik Abels hus, 8th floor
Time:

Workshop at University of Oslo

Time and place: , NHA B1120
Specialization of (stable) birational types is an important tool when studying (stable) rationality in families. A crucial ingredient is to cook up one parameter degenerations such that the limit has certain combinatorial and geometric properties. Nicaise-Ottem studied these questions for hypersurfaces in algebraic tori, and used tropical geometry to construct degenerations that would have been hard (impossible) to construct geometrically. Even after these are constructed one must carefully study the limit in order to apply specialization techniques, this involves both combinatorics and questions about variation of stable birational types. I will talk about the specialization technique in the setup of Nicaise-Ottem, explain some natural questions that appear through the combinatorics, and give some positive results in this direction.
Time:
Time and place: , NHA723

QOMBINE seminar talk by David Jaklitsch (Hamburg)

Time and place: , NHA 1020 and Online
Time and place: , Zoom

Doctoral candidate Alexander Lobbe at the Department of Mathematics will be defending the thesis Machine Learning for the Stochastic Filtering Problem for the degree of Philosophiae Doctor.

Time and place: , Niels Henrik Abels hus, 9th floor

Ingeborg Gjerde (Simula Research Laboratory) presents joint work with Ridgway Scott (University of Chicago).

Abstract: Airflow around airplane wings is characterized by a wide range of flow scales, making it highly challenging to capture numerically. From a simulation viewpoint, the following questions are still being actively investigated: Why do airplanes fly? Can one reliably simulate the lift and drag of an airplane wing? In this talk, I will provide no good answers to these questions. Instead, I want to talk about some interesting results I've stumbled into tangentially, including:
- (Nonlinear) kinetic energy instability analysis, also referred to as Reynolds-Orr instability
- Slip boundary conditions and their connection to D'Alembert's paradox
- Stokes' paradox and its connection to weighted Sobolev spaces. I will show numerical results computed for flow around a cylinder, which serves as a proxy for flow around an airplane wing. In particular, I will talk about the impact of the friction boundary condition on the drag force and flow stability. Finally, I will comment on how these results might be interpreted in view of: New Theory of Flight, J. Hoffman, J. Jansson, C. Johnson (2016), Journal of Mathematical Fluid Mechanics.

Time and place: , NHA B1120
The variety of sums of powers, VSP(F, r) of a homogeneous form F of rank r is the closure in the Hilbert scheme of apolar schemes of length r. A bad limit is a scheme in the closure that is not apolar to F. I will discuss examples of bad limits, including examples for quadrics found by Joachim Jelisiejew that contradicts earlier results on polar simplicies. This is report on work in progress with Jelisiejew and Schreyer and with Grzegorz and Michal Kapustka.
Time and place: , Abels Utsikt (NHA 1259)
Time and place: , Erling Sverdrups plass, Niels Henrik Abels hus, 8th floor