Friday, August 10, 2018 - 8:00am

This year the annual MOPTA conference will be combined with an NSF-TRIPODS sponsored three day summer school for doctoral students and the NSF-DIMACS sponsored workshop on Optimization in Machine Learning which is a part of the DIMACS/Simons Collaboration on Bridging Continuous and Discrete Optimization.

... Read More
Tuesday, May 8, 2018 - 4:00pm

We present a new method of blackbox optimization via gradient sensing with the use of structured random orthogonal matrices, providing more accurate estimators than baselines with provable theoretical guarantees. We show that this algorithm can be successfully applied to learn better quality compressed reinforcement learning policies than these using unstructured... Read More

Thursday, May 3, 2018 - 2:30pm

Schedule of Events Spencer C. Schantz Lecture and 50th Anniversary Symposium (celebrating 50 years of the Ph.D. program) Featuring two talks, a panel discussion and the Spencer C. Schantz Lecture by Mikell Groover Time: 2:30 p.m. - 5:00 p.m. Location: Wood Dining Room, Iacocca Hall, 111 Research Dr, Bethlehem, PAISE Cocktail Reception Time: 5:00 p.m. - 6:00 p.m. ... Read More

Tuesday, April 24, 2018 - 4:00pm

We present a simple restart scheme for speeding-up first order methods for convex optimization problems. Restarts are made based only on achieving specified decreases in objective values, the specified amounts being the same for all optimization problems. Unlike existing restart schemes, the scheme makes no attempt to learn parameter values characterizing the structure... Read More

Tuesday, April 10, 2018 - 4:00pm

The ability of traffic flows to adapt their rate and fairly use all available resources is one of the Internet’s pillars. This traffic characteristic, often referred to as elasticity, has however not been fully considered so far in network routing methodologies. In this work, we propose a new approach to network routing for elastic traffic which models the interaction... Read More

Tuesday, December 5, 2017 - 4:00pm

We propose an unsupervised learning framework for automatically tagging events in basketball game. Our framework uses the the optical player tracking data in the NBA. We first learn the time series of defensive assignment using a novel player and location dependent attraction based model which uses hidden Markov models (HMMs), Gaussian processes, and a``bond breaking... Read More

Thursday, November 30, 2017 - 4:00pm

Cutting planes are the go-to technique for obtaining lower bounds (for minimization problems) in mixed-integer programming and discrete optimization. But a practical weakness is the sequential nature of the method, proceeding from one LP solution to another, step by step by step. As an alternative, we consider the possibility of running separation routines... Read More

Wednesday, November 29, 2017 - 4:00pm
Date: Wednesday, November 29, 2017
Time: 4:00 pm
Title: In Pursuit of the Traveling Salesman: Mathematics at the Limits of Computation.
The traveling salesman problem is easy to state: given a number of cities along with the cost of travel between each pair of them, and the... Read More
Tuesday, November 14, 2017 - 4:00pm

This talk describes a new class of splitting methods for monotone operator problems developed by the author and P. Combettes. It can solve problems involving any finite number of operators and can operate in an asynchronous parallel manner. It has a unique feature for a decomposition algorithm: it does not need to visit every monotone operator at each iteration. The... Read More

Tuesday, November 7, 2017 - 4:00pm

We present a branch-and-cut algorithm for solving discrete bilevel linear programs where the upper-level variables are binary and the lowerlevel variables are either pure integer or pure binary. This algorithm performs local search to find improved bilevel-feasible solutions. We strengthen the relaxed node subproblems in the branch-and-cut search tree by generating... Read More