Conferences, Lectures, & Seminars
Events for September
-
ISE 651-Epstein Institute Seminar Speaker Series
Tue, Sep 06, 2016 @ 03:30 PM - 04:50 PM
Daniel J. Epstein Department of Industrial and Systems Engineering
Conferences, Lectures, & Seminars
Speaker: John Gunnar Carlsson, Ph.D.,
Talk Title: "New problems in modern logistical systems"
Abstract: In recent years, some of the most talked-about developments in the transportation sector include the use of drones, the introduction of last-mile delivery services, and the use of large-scale mapping data. Along with these new developments comes a host of new problems and trade offs. We will discuss three such problems and use the continuous approximation paradigm to reveal basic insights about those factors that influence them most significantly.
Biography: John Gunnar Carlsson is an assistant professor in the Department of Industrial and Systems Engineering at the University of Southern California. He received a Ph.D. in computational mathematics from Stanford University in 2009 and an A.B. in music and mathematics from Harvard College in 2005. He is the recipient of Popular Science magazine's Brilliant 10 Award, the AFOSR Young Investigator Prize, the INFORMS Computing Society (ICS) Prize, and the DARPA Young Faculty Award.
Host: Dr. Jong-Shi Pang
Location: Ethel Percy Andrus Gerontology Center (GER) - 206
Audiences: Everyone Is Invited
Contact: Angela Reneau
-
ISE 651 Epstein Institute Seminar
Tue, Sep 13, 2016 @ 03:30 PM - 04:50 PM
Daniel J. Epstein Department of Industrial and Systems Engineering
Conferences, Lectures, & Seminars
Speaker: Tapas Das, Ph.D., University of South Florida
Talk Title: Upgrading Electric Power Market Infrastructure via Dynamic Pricing and Demand Response
Abstract: Both practitioners and researchers agree that both pricing and demand must play much more proactive roles in better balancing demand of electricity across the hours of a day. A balanced system will reduce the menace of demand and price spikes, routinely experienced by the power networks, and thus reduce the need for expensive reserve generation capacity as well. However, proactive management of pricing and demand would require a more upgraded power market infrastructure than what is currently in place in the U.S. Fortunately, increasing availability of advanced metering and power network infrastructure supported by the Internet of energy IoE will soon pave the way for the desired upgrade. This will facilitate dynamic pricing of electricity by system operators and intelligent demand response by load schedulers (controllers) in smart and connected consumer communities. A dynamic pricing strategy will offer binding prices for each time interval (perhaps, hourly) to the consumer nodes before loads are scheduled. This strategy will replace the current practice of time of use (TOU) pricing. In response to dynamic pricing, the smart communities will optimize their load schedule for all remaining time intervals of the day, as well as manage the use of renewable power generated by the communities.
However, implementing effective dynamic pricing and demand response strategies remains a significant challenge, as models necessary to design such strategies have not been developed and made available for use. This talk outlines the challenge and our approach to address it.
Biography: -“ Tapas K. Das is a professor and chair of the department of Industrial and Management Systems Engineering at the University of South Florida. He is a past chair of the Council of Industrial Engineering Academic Departments Heads (CIEADH), Fellow of IISE, and members of INFORMS and IEEE. His research interest includes policy studies in electric power markets (impact of CO2 emissions control policies on the market, incentive strategies for promoting net zero building, and dynamic pricing and demand response in IoE supported power market) as well as in disease diagnosis and treatment strategies in healthcare delivery.
Host: Dr. Jong-Shi Pang
Location: Ethel Percy Andrus Gerontology Center (GER) - 206
Audiences: Everyone Is Invited
Contact: Angela Reneau
-
Epstein Institute Seminar - ISE 651
Tue, Sep 20, 2016 @ 03:30 PM - 04:50 PM
Daniel J. Epstein Department of Industrial and Systems Engineering
Conferences, Lectures, & Seminars
Speaker: Hongbo Dong, Assistant Professor - Dept. of Mathematics & Statistics - Washington State University
Talk Title: Exploiting Quadratic and Separable Structures in Nonconvex Quadratic Programs via Lift-and-Project
Abstract: In optimization, mixed-integer nonlinear programs MINLP are notoriously difficult to solve to global optimality. It is therefore crucial to exploit problem structures to design effective convex relaxations and/or approximation methods. Same claims hold even when all nonlinear terms are quadratic. We consider a generic sub-structure that comprises a quadratic form and separable (non-convex) constraints. We show how to derive convex relaxations for related non-convex sets in a higher-dimensional space by using conic semidefinite optimization techniques. Essentially by projecting such lifted relaxations back onto the original variable space, we discuss in two concrete scenarios where such lift-and-project techniques improve upon current relaxations, connect with techniques from other areas, and provide new insights. The first scenario concerns generating convex quadratic cutting surfaces to iteratively strengthen classical convex relaxations for mixed-integer quadratic programs. A specialized separation routine (based on coordinate minimization) is developed to avoid (fully) solving semidefinite programs. Our proposed method achieves a more balanced trade-off between strength and computational complexity than existing relaxations, and can be easily incorporated into branch-and-bound algorithms for MINLP. The second scenario concerns the well-known problem of variable selection in statistics and machine learning. We show that lift-and-project methods tightly connect with (folded) concave regularization functions called the Minimax Concave Penalty (MCP) from the statistical community. Our lifting relaxation provides a very different convex relaxation from classical ones (LASSO or l-1 norm) while providing competitive practical performance in certain scenarios
Biography: Hongbo Dong received his Ph.D. in Applied Mathematical and Computational Sciences at the University of Iowa in 2011. After spending two years as a postdoc in a multi-disciplinary optimization group at the University of Wisconsin-Madison, he joined the math department of Washington State University as an Assistant Professor in 2013. His previous research focused on copositive programming, convex relaxations for non-convex problems. Recently he is interested in developing and analyzing novel convex and non-convex formulations for problems in statistics and machine learning. His research results have been published on several optimization and statistical journals including Mathematical Programming, SIAM Journal on Optimization and Biometrika.
Host: Dr. Jong-Shi Pang
Location: 206
Audiences: Everyone Is Invited
Contact: Angela Reneau
-
Epstein Institute Seminar - ISE 651
Tue, Sep 27, 2016 @ 03:30 PM - 04:50 PM
Daniel J. Epstein Department of Industrial and Systems Engineering
Conferences, Lectures, & Seminars
Speaker: Dr. Michael L. Overton, Professor of Computer Science and Mathematics at the Courant Instittute of Mathematical Sciences, New York University
Talk Title: Nonsmooth, Nonconvex Optimization: Algorithms and Examples
Abstract: In many applications one wishes to minimize an objective function that is not convex and is not differentiable at its minimizers. We discuss two algorithms for minimization of nonsmooth, nonconvex functions. Gradient Sampling is a simple method that, although computationally intensive, has a nice convergence theory. The method is robust and the convergence theory has recently been extended to constrained problems.
BFGS is a well-known method, developed for smooth problems, but which is remarkably effective for nonsmooth problems too. Although our theoretical results in the nonsmooth case are quite limited, we have made some remarkable empirical observations and have had broad success in applications. Limited Memory BFGS is a popular extension for large problems, and it is also applicable to the nonsmooth case, although our experience with it is more mixed. Throughout the talk we illustrate the ideas through examples, some very easy and some very challenging. Our work is with Jim Burke U. Washington and Adrian Lewis Cornell.
Biography: Michael L. Overton is Professor of Computer Science and Mathematics at the Courant Institute of Mathematical Sciences, New York University. He received his Ph.D. in Computer Science from Stanford University in 1979. He is a fellow of SIAM Society for Industrial and Applied Mathematics and of the IMA -Institute of Mathematics and its Applications, UK. He served on the Council and Board of Trustees of SIAM from 1991 to 2005, including a term as Chair of the Board from 2004 to 2005. He served as Editor-in-Chief of SIAM Journal on Optimization from 1995 to 1999 and of the IMA Journal of Numerical Analysis from 2007 to 2008, and was the Editor-in-Chief of the MPS Mathematical Programming Society-SIAM joint book series from 2003 to 2007. He is currently an editor of SIAM Journal on Matrix Analysis and Applications, IMA Journal of Numerical Analysis, Foundations of Computationa Mathematics, and Numerische Mathematik. His research interests are at the interface of optimization and linear algebra, especially nonsmooth optimization problems involving eigenvalues, pseudospectra, stability and robust control. He is the author of Numerical Computing with IEEE Floating Point Arithmetic SIAM, 2001.
Host: Dr. Jong-Shi Pang
Location: 206
Audiences: Everyone Is Invited
Contact: Angela Reneau