Conference: Statistical Optimal Transport

Statistical Optimal Transport

July 24-25, 2018

 The aim of this workshop is to present recent advances at the cutting edge of the theory of optimal transport and modern nonparametric statistics and to discuss tje cooperation between MIT and Skoltech on this topic.

Venue: The conference will be held at the Skolkovo Institute of Science and Technology (Moscow, Nobelya Ulitsa 3).

Registration: The number of participants is limited. Any person wishing to participate must register at the following link: Registration

Program: A preliminary version will be available soon.

Mini-courses (each course comprises two one-hour lectures):

Alexander Kolesnikov (National Research University Higher School of Economics), Mass transportation: the classical Monge-Kantorovich problem and recent developments 
Thibaut Le Gouic (Ecole Centrale de Marseille), Curvature and barycentre: application to Wasserstein space

List of participants:

Aden Forrow (MIT) 
Alexander Kolesnikov (HSE) 
Alexey Kroshnin (IITP RAS) 
Thibaut Le Gouic (Ecole Centrale de Marseille) 
Alexey Naumov (HSE) 
Quentin Paris (HSE) 
Philippe Rigollet (MIT) 
Vladimir Spokoiny (WIAS, Skoltech, HSE) 
Alexandra Suvorikova (WIAS) 
Jonathan Weed (MIT) 
to be continued…

Program committee: Philippe Rigollet and Vladimir Spokoiny;

Local organizing committee: Alexander Kolesnikov, Vladimir Spokoiny, Alexandra Suvorikova

The workshop is organized by the 
Skolkovo Institute of Science and Technology

To learn more about this event and the Structural Learning Group, click here

Program of events for Tuesday, 24 July: 

       
Time   Speaker Title
9:50 -10:00 Opening    
10:00 – 11:10 Minicourse 1 Alexander Kolesnikov, National Research University Higher School of Economics

Mass transportation: the classical Monge-Kantorovich problem and recent developments

11:10 – 11:40 Coffee break    
11:40 -12:50 Invited talks    
11:40 – 12:15   Eugene Stepanov, St. Petersburg Branch of the Steklov Mathematical Institute

Quantization/optimal location for signed measures

12:15 – 12:50   Jonathan Weed, Massachusetts Institute of Technology

Minimax Wasserstein estimation of smooth densities.

12:50 – 14:20 Lunch    
14:20 – 15:30 Minicourse 2 Thibaut Le Gouic, Ecole Centrale de Marseille

Curvature and barycentre: application to Wasserstein space

15:30 – 16:00 Coffee break    
16:00 – 17:30 Invited talks    
16:00-16:30   Sebastian Melf Boeckel, Humboldt University Berlin

Nonparametric multivariate test theory

16:30 – 17:00   Nikita Gladkov, National Research University Higher School of Economics

Multidimensional transport problem

17:00 – 17:30   Alexaey Kroshnin, Institute for Information Transmission Problems

Central limit theorem in 2-Wasserstein space

17:30 – 20:00 Welcome party    

Program of events for Wednesday, 25 July: 

Time   Speaker Title
10:00 – 11:10 Minicourse 1 Alexander Kolesnikov, National Research University Higher School of Economics

Mass transportation: the classical Monge-Kantorovich problem and recent developments

11:10 – 11:40 Coffee break    
11:40 -12:50 Invited talks    
11:40 – 12:15   Franz Besold, Humboldt University Berlin

Adaptive Nonparametric Clustering

12:15 – 12:50   Evgeny Burnaev, Skolkovo Institute of Science and Technology

Anonymous walk embeddings

12:50 – 14:20 Lunch    
14:20 – 15:30 Minicourse 2 Thibaut Le Gouic, Ecole Centrale de Marseille

Curvature and barycentre: application to Wasserstein space

15:30 – 16:00 Coffee break    
16:00 – 17:00 Invited talks    
16:00-16:30   Pavel Dvurechensky, Weierstrass Institute for Applied Analysis and Stochastics

Computational Optimal Transport: Accelerated Gradient Descent vs Sinkhorn’s Algorithm

16:30-17:00   Darina Dvinskikh, Moscow Institute of Physics and Technology

Stochastic distributed computation of Wasserstein barycenter

sk-logo

Tweet about this on Twitter0Share on Facebook3Pin on Pinterest0Share on Tumblr0Share on VK