Operator Splitting Methods in Data Analysis

Workshop

When

March 20, 2019
8 a.m.–6 p.m. ET
March 21, 2019
All Day
March 22, 2019
8 a.m.–6 p.m. ET

Organized by

Patrick L. Combettes Christian L. Müller
Distinguished Professor, Project Leader in Computational
Mathematics, North Carolina Statistics, Numerical Analysis,
State University CCM, Flatiron Institute

Red, curved line with many arrows pointing to it. Next to it says

The 2nd workshop on Operator Splitting Methods in Data Analysis brings together experts from nonlinear analysis, optimization, data analysis, and statistics to advance the field of operator splitting methods and its application to statistics and data analysis.

Specific topics include:
(i) general forms of equilibria that are captured by monotone inclusions,
(ii) nonconvex/nonmonotone splitting methods
(iii) scaling up of splitting strategies as the numbers of observations and features grow
(iv) robustness of splitting methods under perturbations and modeling errors
(v) New and nonstandard variational formulations for data analysis problems
(vi) Applications of splitting methods in data analysis and statistics

A unique aspect will be again the strong focus on presentations that illuminate technical difficulties and highlight open problems.

Information about the 2018 edition of this workshop, held at SAMSI, can be found here (https://www.samsi.info/programs-and-activities/research-workshops/operator-splitting-methods-data-analysis/).

AGENDA

For a more detailed schedule go here.

Wednesday, March 20

8:15AM – 9:00AM	Breakfast
9:00AM – 9:15AM	Welcome remarks by Patrick L. Combettes and Christian L. Müller
9:15AM – 10:45AM	3 25-minute Talks
9:15AM – 9:30AM	Saverio Salzo
9:45AM – 10:15AM	Ivan Selesnick
10:15AM – 10:45AM	Noah Simon
10:45AM – 11:15AM	Break
11:15AM – 12:15PM	2 25-minute Talks
11:15AM – 11:45AM	Sylvain Sorin
11:45AM – 12:15PM	Sebastian Stich
12:15PM – 1:30PM	Lunch
1:30PM – 2:30PM	2 25-minute Talks
1:30PM – 2:00PM	Weijie Su
2:00PM – 2:30PM	Ryan Tibshirani
2:30PM – 3:00PM	Break
3:00PM – 4:30PM	3 25-minute Talks
3:00PM – 3:30PM	Charles Epstein
3:30PM – 4:00PM	Isao Yamada
4:00PM – 4:30PM	Maicon Marques Alves
4:30PM – 5:00PM	Break
5:00PM – 6:00PM	2 25-minute Talks
5:00PM – 5:30PM	Hedy Attouch
5:30PM – 6:00PM	Jacob Bien
6:00PM – 8:30PM	Reception

Thursday, March 21

8:30AM – 9:00AM	Breakfast
9:00AM – 10:30AM	3 25-minute Talks
9:00Am – 9:30AM	Radu Bot
9:30AM – 10:00AM	Luis Briceño-Arias
10:00AM – 10:30AM	Eric Chi
10:30AM – 11:00AM	Break
11:00AM – 12:00PM	2 25-minute Talks
11:00AM – 11:30AM	Jonathan Eckstein
11:30AM – 12:00PM	Aurelie Lozano
12:00PM – 1:30PM	Lunch
1:30PM – 2:30PM	2 25-minute Talks
1:30PM – 2:00PM	Aditya Mishra
2:00PM – 2:30PM	Walaa Moursi
2:30PM – 3:00PM	Break
3:00PM – 4:30PM	3 25-minute Talks
3:00PM – 3:30PM	Marcello Pelillo
3:30PM – 4:00PM	Teemu Pennanen
4:00PM – 4:30PM	Massimiliano Pontil
4:30PM – 5:00PM	Break
5:00PM – 6:00PM	2 25-minute Talks
5:00PM – 5:30PM	Vianney Perchet
5:30PM – 6:00PM	Marcelo Pereya
6:00PM – 9:00PM	Dinner at 6:30PM

Friday, March 22

8:30AM – 9:00AM	Breakfast
9:00AM – 10:30AM	3 25-minute Talks
9:00AM – 9:30AM	Jean-Christophe Pesquet
9:30AM – 10:00AM	Aaditya Rangan
10:00AM – 10:30AM	Andreas Buja
10:30AM – 11:00AM	Break
11:00AM – 11:30AM	1 25-minute Talk
11:00AM – 11:30AM	Terry Rockafellar
11:30AM – 12:30PM	Summary of the Workshop / Discussion
12:30PM – 1:00PM	Lunch

PARTICIPANTS

Participant	Affiliation	Title (Link to Slides)
Maicon Marques Alves	Federal University of Santa Catarina	Inexact relative-error proximal point algorithms
Hedy Attouch	IMAG, Université de Montpellier, CNRS	Fast optimization through inertial dynamics with Hessian Driven Damping
Jacob Bien	University of Southern California	Optimization Along the Path
Radu Boț	Universität Wien	Four (plus one) open questions about the minimization of structured nonsmooth and nonconvex functions
Luis Briceño-Arias	Universidad Técnica Federico Santa Maria	On projecting in primal-dual splitting methods for solving constrained convex optimization
Andreas Buja	University of Pennsylvania	A Probability Coverage Problem that Could Solve the Reproducibility Crisis in the Empirical Sciences
Eric Chi	North Carolina State University	Better Tools for Proving Convergence for MM Algorithms?
Jonathan Eckstein	Rutgers University
Charles Epstein	University of Pennsylvania	The Geometry of the Phase Retrieval Problem
Aurelie Lozano	IBM T.J. Watson Research Center
Walaa Moursi	University of Waterloo	The averagedness of Douglas-Rachford and forward-backward operators
Marcello Pelillo	University of Venice
Teemu Pennanen	King’s College London	Information based complexity in stochastic optimization
Massimiliano Pontil	Istituto Italiano di Tecnologia (IIT)	Empirical Risk Minimization Under Fairness Constraints
Vianney Perchet	Ecole Normale Supérieure Paris-Saclay & Criteo AI Lab	Private Learning/Optim
Marcelo Pereyra	Heriot-Watt University	Proximal algorithms with mysterious “plug-and-play” operators
Jean-Christophe Pesquet	Université Paris-Saclay	Deep Neural Networks and Variational Inequalities
Aaditya Rangan	Courant Institute, NYU
Terry Rockafellar	University of Washington	Problem Decomposition in Nonconvex Optimization
Saverio Salzo	Istituto Italiano di Tecnologia	Improving the Random Block-Coordinate Forward-Backward Method
Ivan Selesnik	New York University
Noah Simon	University of Washington	Stochastic Gradient Methods for non-parametric estimation
Sylvain Sorin	Sorbonne Université	No-Regret criteria in learning, games and convex optimization
Sebastian Stich	EPFL, Switzerland	Communication Efficient SGD for Distributed Optimization
Weijie Su	University of Pennsylvania	Taming the Devil of Gradient-based Optimization with the Angel of Differential Equations
Ryan Tibshirani	Carnegie Mellon University	Dykstra’s Algorithm, ADMM, and Coordinate Descent: Connections, Insights, and Extensions
Isao Yamada	Tokyo Institute of Technology	A Proximal Splitting Algorithm for Linearly Involved Generalized Minimax Concave Penalty Models