P.C. Rossin College of
Engineering and Applied Science

Upcoming Events

ISE SEMINAR SERIES

Title: Learning with difference-of-convex sparsity functions

Speaker: Miju Ahn, Southern Methodist University

Time & Location:

Tuesday, October 8, 2019
4:30-5:30pm
Mohler Lab, #453

Abstract      

Variable selection is a fundamental methodology in statistical learning performed to obtain sparse representation of the trained model parameters. The first part of this talk emphasizes on studying properties of stationary solutions of the learning formulations where sparsity functions are used in the objective function. We derive bounds for the distance between the stationary solutions and a vector that is possibly the ground truth, and the model predictions produced by them. Furthermore, the inclusion relationships between their support sets are studied. Some of the results are generalization of the existing theory shown for the L1-norm regularized least squares minimization problem. The second part emphasizes on the constraint system defined by the sparsity functions. A system of affine sparsity constraints (ASCs) formulates various logical conditions among the variables as hard constraints. Structural properties of the set including its closeness and the characterization of its closure are analyzed. The set convergence of the approximation sets to the set of ASCs is presented.

Bio sketch

Miju Ahn is an Assistant Professor in the Department of Engineering Management, Information, and Systems (EMIS) of the Lyle School of Engineering at the Southern Methodist University. She received her Ph.D. in Industrial and Systems Engineering from University of Southern California. For undergraduate studies, she received B.A. in Applied Mathematics and obtained Industrial Engineering and Operations Research (IEOR) minor at University of California, Berkeley. Her research interests include computational optimization, operations research, statistical/machine learning, and big-data science.

 


Past Events

Fall 2019
 
Tuesday, September 17, 2019 - 4:30pm
Ramteen Sioshansi, The Ohio State University
"Can We Get Market and Regulatory Designs ‘Right’ for Energy Storage?"
 
Tuesday, September 10, 2019 - 4:30pm
ISE Seminar Series - Adam Elmachtoub, Columbia University
"Smart "Predict, then Optimize""
 
Spring 2019
 
Tuesday, May 7, 2019 - 4:00pm
ISE Seminar Series - Alec Koppel, U.S. Army Research Laboratory
"Policy Search for Reinforcement Learning in Continuous Spaces: Improved Limits and Reduced Variance"
 
Tuesday, April 9, 2019 - 4:00pm
ISE Seminar Series - Aida Khajavirad (Rutgers & NYU)
"Novel Polyhedral Relaxations for Mixed-Integer Polynomial Optimization Problems"
 
Tuesday, April 2, 2018 - 4:00pm
INFORMS Chapter Distinguished Speaker Series - Stefan Wild, Argonne National Laboratory
"Exploiting Structure in Derivative-Free Nonlinear Optimization to Advance Science and Engineering"
 
Tuesday, March 26, 2019 - 4:00pm
ISE Seminar Series - Clément Royer, University of Wisconsin-Madison
"Nonconvex Optimization via Newton-CG Methods with Complexity Guarantees"
 
Tuesday, March 19, 2019 - 4:00pm
ISE Seminar Series - William K. Klimack, Chevron
"Decision Analysis at Chevron"
 
Tuesday, March 5, 2019 - 4:00pm
ISE Seminar Series - James Luedtke, University of Wisconsin-Madison
"Optimizing Truck Dispatching Decisions in Open-pit Mining using Integer Programming"
 
Tuesday, February 12, 2019 - 4:00pm
ISE Seminar Series - J. Cole Smith, Clemson University
"The Maximum Semicontinuous Flow Problem"
 
Tuesday, January 29, 2019 - 4:00pm
ISE Seminar Series - Giacomo Nannicini, IBM T.J. Watson Research Center
"Fully Polynomial-Time Approximation Schemes for Stochastic Dynamic Programs: Theory and Applications"
 
Fall 2018
 
Tuesday, December 4, 2018 - 4:00pm
ISE Seminar Series - Andreas Wächter, Northwestern University
"Nonlinear Programming Formulations of Chance-Constraints"
 
Tuesday, November 27, 2018 - 4:00pm
INFORMS Chapter Distinguished Speaker Series - Gerard Cornuejols, Carnegie Mellon University
"Packing and Covering"
 
Tuesday, November 13, 2018 - 4:00pm
ISE Seminar Series - Mert Gürbüzbalaban, Rutgers University
"Momentum Acceleration Under Random Gradient Noise: From Convex to Non-Convex Optimization"
 
Tuesday, October 30, 2018 - 4:00pm
ISE Seminar Series - Michael Katehakis, Rutgers University
"Reinforcement Learning: Connections Between MDPs and MAB Problems"
 
Tuesday, October 2, 2018 - 4:00pm
ISE Seminar Series - Michael L. Overton, Courant Institute of Mathematical Sciences, NYU
"Nonsmooth, Nonconvex Optimization: Algorithms and Examples"
 
Spring 2018
 
Tuesday, May 8, 2018 - 4:00pm
ISE Seminar Series - Krzysztof Choromanski, Google
"Structure is all that you need - Learning Compressed RL Policies Via Orthogonal Gradient Sensing"
 
Tuesday, May 1, 2018 - 4:00pm
ISE Seminar Series - Henry Lam, Columbia University
"Assessing Solution Quality in Stochastic Optimization with Limited Data"
 
Tuesday, April 24, 2018 - 4:00pm
INFORMS Chapter Distinguished Speaker Series - James Renegar, Cornell University
"A Simple Nearly-Optimal Restart Scheme for Speeding-Up First Order Methods"
 
Thursday, April 19, 2018 - 4:00pm
ISE Seminar Series - Matthias Walter, RWTH Aachen University
"Investigating Polyhedra by Oracles"
 
Tuesday, April 10, 2018 - 4:00pm
ISE Seminar Series - Stefano Coniglio, University of Southampton
"Network Routing Through the Internet as a Stackelberg Game"
 
Tuesday, March 6, 2018 - 4:00pm
ISE Seminar Series - Bo Zeng, University of Pittsburgh
"Robust and Chance Constrained Optimization in Power and Logistics Systems"
 
Fall 2017
 
Tuesday, December 5, 2017 - 4:00pm
ISE Seminar Series - Garud Iyengar, Columbia University
"Automatic Event Detection in Basketball"
 
Thursday, November 30, 2017 - 4:00pm
Spencer C. Schantz Technical Lecture - William J. Cook, University of Waterloo
"Guiding the Cutting-Plane Method"
 
Wednesday, November 29, 2017 - 4:00pm
Spencer C. Schantz Distinguished Lecture Series Presents William J. Cook
"In Pursuit of the Traveling Salesman: Mathematics at the Limits of Computation"
 
Tuesday, November 14, 2017 - 4:00pm
ISE Seminar Series - Jonathan Eckstein, Rutgers University
"Asynchronous Projective Splitting for Convex Optimization and Monotone Inclusion Problems"
 
Tuesday, November 7, 2017 - 4:00pm
ISE Seminar Series - Osman Y. Ozaltin, North Carolina State University
"A Branch-and-cut Algorithm for Discrete Bilevel Linear Programs"
 
Tuesday, October 31, 2017 - 4:00pm
ISE Seminar Series - Oleg Prokopyev, University of Pittsburgh
"Sequential Max-Min Bilevel Linear Programming with Incomplete Information and Learning"
 
Tuesday, October 10, 2017 - 4:00pm
ISE Seminar Series - Ion Necoara, University Politehnica of Bucharest
"Conditions for Linear Convergence of (Stochastic) First Order Methods"
 
Thursday, October 5, 2017 - 4:00pm
INFORMS Chapter Distinguished Speaker Series: Nick Trefethen, University of Oxford
 
Tuesday, September 26, 2017 - 4:00pm
ISE Seminar Series - Andrew Dorsett, Wolfram Research
"Mathematica 11 in Education and Research"
 
Tuesday, September 19, 2017 - 4:00pm
ISE Seminar Series - Alejandro Ribeiro, University of Pennsylvania
"High Order Methods in Empirical Risk Minimization"
 
Tuesday, August 22, 2017 - 4:00pm
ISE Seminar Series - Giorgio Fasano, Thales Alenia Space
"Solving Non-Standard Packing Problems by Global Optimization and Heuristics"