2010 Seminars

Please note: The seminar this week will be composed of two short talks.

Friday, December 3, 2010

2:30 p.m. - 3:30 p.m.

Mohler Lab Room 451


Talk #1
Amr El-Bakry

Title: Optimization Challenges in the Oil and Gas Industry

ABSTRACT

The talk is a high-level overview of the journey that a hydrocarbon molecule takes from the reservoir to the consumer. Throughout the talk, the speaker will point out where optimization opportunities arise and discuss the challenges optimization technology faces in addressing those problems.

BIOGRAPHY

Amr El-Bakry is a Senior Research Associate and an Optimization Advisor at Corporate Strategic Research, ExxonMobil Research and Engineering Company. He received his PhD in Computational and Applied Mathematics from Rice University in 1991. He worked in Academia for 7 years before joining ExxonMobil Upstream Research Co. in 1998. He worked in optimization projects covering imaging, drilling, production, development, and logistics. In 2008 he moved to Strategic Corporate Research. His research interests cover the areas of computational optimization and decision modelling in the presence of uncertainty. Amr is an editor in the Journal of Optimization and Engineering.


Talk #2
Jin-Hwa Song

Title: A Practical Approach to a Maritime Inventory Routing Problem

ABSTRACT

ExxonMobil transports significant volumes of vacuum gas oil (VGO) on an annual basis from supply points in Northwest Europe to U.S. refineries. Optimizing these transportation costs via modern mathematical programming technology allows for significant cost savings. We introduce a practical problem for simultaneous optimization of ship routing and inventory management of a bulk refinery product. Even though this ship inventory routing problem and the conventional Inventory Routing Problem (IRP) have similar structures, differences arise in various characteristics such as complex routing and time dependent costs, cargo draft limits at ports and allowing routes with multiple pick-ups and drop-offs. We develop discrete time optimization models and practical heuristic algorithms which address these various real-world issues. This presentation is based on the joint work with Kevin Furman from ExxonMobil Upstream Research Company.

BIOGRAPHY

Jin-Hwa Song is currently leading optimization research efforts at Corporate Strategic Research, ExxonMobil Research and Engineering Company. He received his PhD from the H. Milton Stewart School of Industrial and Systems Engineering at the Georgia Institute of Technology. Since joining ExxonMobil, he has been working on various optimization and decision science related projects, as well as science build research programs with several universities.


Professor Matthew Bailey

Title: Eliciting Patients’ Revealed Preferences: An Inverse MDP Approach

Bucknell University School of Management and Adjunct Faculty Research Investigator at Geisinger Health System

Friday, November 19, 2010

2:30 p.m. - 3:30 p.m.

Mohler Lab Room 451

ABSTRACT

Eliciting patient’s preferences over various health states is an important problem in health care decision modeling. Direct approaches, which involve asking patients various abstract questions, have significant drawbacks. We propose a new approach that infers patient preferences based on observed decisions via inverse Markov decision process techniques. We consider the specific case of the timing of a living-donor liver transplantation as an illustration.

BIOGRAPHY

Dr. Bailey holds a joint appointment as an Assistant Professor at Bucknell University in the School of Management and an Adjunct Faculty Research Investigator at Geisinger Health System. His primary research interests are in sequential decision-making under uncertainty and operations research applied to healthcare. At Geisinger, his time is balanced between medical decision making in the Center for Health Research and hospital operations within Innovations. He has published in journals such as Operations Research, IIE Transactions, and Naval Research Logistics. Dr. Bailey was previously on the faculty at the University of Pittsburgh’s Industrial Engineering Department and affiliated with the Medical School's Section of Decision Sciences and Clinical Systems Modeling.


Professor John E. Mitchell

Title: Finding Global Optima of Mathematical Programs With Complementarity Constraints

Rensselaer Polytechnic Institute, NY

Friday, November 5, 2010

2:30 p.m. - 3:30 p.m.

Mohler Lab Room 451

ABSTRACT

Mathematical programs with complementarity constraints constitute a powerful modeling paradigm. Applications include inverse optimization problems, hierarchical optimization problems, and Stackelberg Games. In this talk, two algorithms for solving linear programs with complementarity constraints are presented: a logical Benders decomposition approach and a branch-and-cut method. Classes of cutting planes and surfaces are described. Extensions of the algorithms to convex quadratic programs with complementarity constraints are also briefly discussed.

BIOGRAPHY

John E. Mitchell is a Professor of Mathematical Sciences at Rensselaer Polytechnic Institute, with a joint appointment in Industrial and Systems Engineering. He received his PhD from Cornell University, working with Professor Michael Todd on interior point cutting plane methods. He has published papers on interior point methods, polyhedral theory for integer programming, semidefinite and cone programming, mathematical programming problems with complementarity constraints, and in application areas including infrastructure restoration, sports scheduling, robot motion planning, and sensor location.


Professor Jiming Peng

Title: Sparse Solutions to Standard Quadratic Optimization Problems With Random Data

University of Illinois at Urbana-Champaign

Friday, October 29, 2010

2:30 p.m. - 3:30 p.m.

Mohler Lab Room 451

ABSTRACT

Sparse solutions in optimization has been a major concern for optimization problems from various applications such as image processing and portfolio selection. It is well known that several classes of linear optimization problems such as the linear assignment problem, transportation problem and the L_1 minimization problem with linear equality constraints have sparse solutions. It has also been observed in experiments for long that several classes of quadratic optimization problems (QP) such as Markowitz's mean-variance model for portfolio selection always have sparse optimal solutions. However, so far little is known from a theoretical perspective.

In this talk, we present a new theoretical framework to interpret why certain classes of QPs do have sparse solutions. For this, we first use the optimality conditions for the so-called standard quadratic optimization problem to establish an intrinsic relation between the sparsity of the optimal solution of the underlying QP and some probability events. Then we show that with a very high probability, the underlying QP has a very sparse solution if the input data of the associated QP follows certain distributions such as uniform, exponential and normal distributions. The new theoretical analysis also sheds new light on the design of effective algorithm for classes of QPs by exploring the sparsity in the global optimal solution of the problem. If time allows, we shall also discuss some extensions and open questions.

BIOGRAPHY

Jiming Peng is an Assistant Professor in the Department of Industrial and Enterprise System Engineering, University of Illinois at Urbana-Champaign. He received his PhD degree from Delft University of Technology, the Netherlands in 2001. From 2001 to 2006, he worked in the department of computing and software, McMaster University, Canada.

Professor Peng’s research interest covers several areas in the field of mathematical programming including numerical algorithms for variational inequalities and complementarity problems, interior-point methods for linear conic programming, approximation to NP-hard problems and applications in data mining, computer vision and engineering design. He has published dozens of peer-reviewed papers in major optimization journals, a research monograph and more than a dozen papers in highly ranked CS/IEEE conferences. Dr. Peng has been an Associate Editor for the Springer journal Optimization Letters since 2006.


Dr. Michael Kuby

Title: Optimal Location of Refueling Stations for the Transition to Alternative-Fuel Vehicles

School of Geographical Sciences and Urban Planning, Arizona State University

Friday, October 8, 2010

2:30 p.m. - 3:30 p.m.

Mohler Lab Room 451

ABSTRACT

It is widely acknowledged that the greatest barrier to the transition away from petroleum-based transportation is the lack of refueling station infrastructure. As such, it is important to deploy the early stations as efficiently as possible. This paper presents a model, case study, and several extensions for optimal location of refueling stations for alternative-fuels, such as hydrogen, ethanol, biodiesel, natural gas, or electricity. It begins with some background on alternative fuels and a brief literature review of competing optimization approaches. It then introduces the Flow-Refueling Location Model (FRLM), which locates p stations to maximize the flows or paths that can be refueled given a maximum driving range of vehicles. Because of the driving range limitation, longer paths require combinations of stations to refuel round-trip travel. A mixed-integer linear programming (MILP) formulation of the model is presented, as well as greedy and genetic heuristic algorithms. The heuristic algorithms are applied to locate hydrogen-refueling stations at the state and metropolitan scales in Florida. The paper concludes with several extensions of the basic FRLM.

BIOGRAPHY

Michael Kuby received a Bachelor’s degree from The University of Chicago in 1980 and a Ph.D. from Boston University in 1988, both in Geography. He is a Professor in the School of Geographical Sciences and Urban Planning at Arizona State University, where he has taught since 1988. His research centers on creating optimization models for facility location or transport network design, mainly for sustainable energy and transport systems. His work with the World Bank on energy and railway transport in China was a Finalist for the 1994 Franz Edelman Award for Management Science Achievement. He has published in journals such as Interfaces, European Journal of Operational Research, International Journal of Hydrogen Energy, Socio-Economic Planning Sciences, Geographical Analysis, Advances in Water Resources, Energy Policy, and Transportation Research-A. Dr. Kuby is on the editorial board of the Journal of Transport Geography, and is currently Area Editor for Location Science for Networks and Spatial Economics. He is co-author of Human Geography in Action, an interactive textbook/lab manual published by John Wiley & Sons, now in its 5th edition.


Dr. Florian Potra, Ph.D.

Title: Interior Point Methods for Protein Image Alignment

Professor of Mathematics and Statistics, University of Maryland, Baltimore County

Friday, October 1, 2010

2:30 p.m. - 3:30 p.m.

Mohler Lab Room 451

ABSTRACT

One of the core technologies for obtaining protein mixtures is provided by two dimensional polyacrylamide gel electrophoresis. In order to analyze variations in the protein gels obtained from different groups that account for biological variations we must first eliminate distortions to properly align images. The image alignment is recognized as a major bottleneck in proteomics. We formulate the image alignment problem as a large-scale quadratic programming problem, which can be solved in polynomial time by interior point methods. Numerical results illustrate the effectiveness of this approach.

BIOGRAPHY

Florian Potra earned a Ph.D. in Mathematics from the University of Bucharest, Romania. After an Andrew Mellon Postdoctoral Fellowship at the University of Pittsburgh, he joined the faculty of the University of Iowa, first as an Associate Professor of Mathematics, and then as a Professor of Mathematics and Computer Science. Between 1997-1998, he served as a Program Director in Applied and Computational Mathematics at the National Science Foundation. Since 1998, he has been a Professor of Mathematics and Statistics at the University of Maryland Baltimore County. He is also a Faculty Appointee at the Mathematical and Computational Sciences Division of The National Institute of Standards and Technology. Dr. Potra has published over 120 research papers in prestigious professional journals. He is the Regional Editor for the Americas of the journal "Optimization Methods and Software", and serves on the editorial board of three other well-known mathematical journals.


Dr. Alice E. Smith

Title: Retail Facilities Design - Considering Revenue, Adjacencies and Aisle Structure

Department of Industrial and Systems Engineering , Auburn University

Friday, September 10, 2010

2:30 p.m. - 3:30 p.m.

Mohler Lab Room 451

ABSTRACT

In this presentation, a model and a solution approach for the spatial design of retail stores is proposed. Unlike previous models that almost exclusively focus on shelf space allocation, this research focuses on block layout of the entire store area. Ideas from both shelf space allocation and manufacturing facility layout are used in formulating the model. The model considers the area allocated to departments, exposure of departments to the aisle structure, and adjacency requirements (all related to revenue generation) among departments subject to spatial constraints imposed on the departments. A basic constraint on the layout is that the aisle structure of the store needs to be in the form of a racetrack, and the racetrack aisle is handled essentially as a department, with variable area. Due to the complex nature of the model, a two stage optimization algorithm is devised to solve the problem. In the first stage, non-linear optimization is used to optimize the area allotments of the departments and the aisle area. In the second stage, a tabu search algorithm is used to optimize the location of the departments within the store, maximize the normalized adjacency score, maintain the aisle width within specified limits and maintain aspect ratio constraint for the departments. Several test cases are presented. This study is one of the first efforts to formulate and solve the spatial design of retail stores.

BIOGRAPHY

Alice E. Smith is Professor and Chair of the Industrial and Systems Engineering Department at Auburn University. Previous to this position, she was on the faculty of the Department of Industrial Engineering at the University of Pittsburgh, which she joined in 1991 after industrial experience with Southwestern Bell Corporation. Dr. Smith has degrees in engineering and business from Rice University, Saint Louis University and Missouri University of Science and Technology. Dr. Smith was awarded the INFORMS WORMS Award for the Advancement of Women in OR/MS in 2009. She was named an Auburn University Philpott-WestPoint Stevens Distinguished Professor in 2001, received the Senior Research Award of the College of Engineering at Auburn University in 2001 and the University of Pittsburgh School of Engineering Board of Visitors Faculty Award in 1996.


John T. BettsDr. John T. Betts
The Boeing Company

Public Lecture - Friday, April 16, 2010
2:30 p.m. - 3:30 p.m. Perella Auditorium, Rauch Business Center

Title: What Does a Rocket Scientist Really Do?

When describing something dreadfully simple or blatantly obvious we often proclaim "It is Not Rocket Science." Yet while we may readily agree that "rocket science" is not a synonym for "simple," it sheds no light on what a rocket scientist really does. In this talk I will reflect on the many subjects encountered during a career in the aerospace industry. Hopefully, I can convey a sense of what rocket science is all about!


Technical Talk - Thursday, April 15, 2010
2:30 p.m. - 3:30 p.m. Room 451, Mohler Lab

Title: Algorithmic Choices When Solving an Optimal Control Problem

When designing a computational algorithm for solving a complicated problem it is usually necessary to choose between one or more alternatives. Should functions be approximated using polynomials, rational functions, or Fourier series? Should a linear system be solved using Gaussian elimination or orthogonal decomposition? Ultimately the design of the algorithm must select a combination of techniques that are efficient, robust, implementable and that have a sound theoretical basis. This talk will review many of the choices needed to construct an effective method for solving an optimal control problem.

Dr. Betts's simulations from MOPTA 2002 conference. This simulation shows how long it would take a rocket to reach the moon from the earth using a low energy force (such as the air from a hair dryer).

Biography

John T. Betts received a B.A. degree from Grinnell College in 1965 with a major in physics and minor in mathematics. He attended graduate school at Purdue University and in 1967 received an M.S. in Astronautics with a major in orbit mechanics. He received a Ph.D. in aeronautical engineering from Purdue in 1970, specializing in optimal control theory. He joined The Aerospace Corporation in 1970 as a Member of the Technical Staff, and from 1977-1987 was manager of the Optimization Techniques Section of the Performance Analysis Department. He joined the Boeing Company, serving as manager of the Operations Research Group of Boeing Computer Services from 1987-1989. He served as a Technical Fellow in the Mathematics and Computing Technology Division, until his retirement in 2009, during which time he provided technical support to all areas of the Boeing Co. Dr. Betts is a member of AIAA and SIAM with active research in nonlinear programming and optimal control theory. In 2004, he was granted an "outstanding aerospace engineer award" by Purdue University. He has over 50 technical publications, and is the author of two books on optimal control methods.

About Spencer C. Schantz

This lecture series is endowed in the name of the late Spencer C. Schantz, who graduated from Lehigh in 1955 with a B.S. in Industrial Engineering. Following progressive responsibilities with several electrical manufacturing companies, in 1969 he founded U.S. Controls Corporation and became its first CEO and President. The Spencer C. Schantz Distinguished Lecture Series was established by his wife Jerelyn as a valuable educational experience for faculty, students and friends of Lehigh’s Industrial and Systems Engineering department.


Dr. Margaret H. Wright, Ph.D.
Margaret Wright

Title: Optimization Without Derivatives: Consensus and Controversies

Courant Institute of Mathematical Sciences, New York University

Friday, April 9, 2010

2:30 p.m. - 3:35 p.m.

Mohler Lab Room 451

ABSTRACT

Non-derivative methods for optimization have had a sometimes rocky relationship for more than 50 years with applied mathematicians who specialize in optimization. Although practitioners have never wavered in their fondness for non-derivative methods, their mathematical foundations were mostly lacking until the late 1980s. Since then, significant progress has been made concerning theoretical underpinnings, but several perplexing mysteries remain. In addition, there has been continuing and lively controversy about which methods are “most effective'' on real-world applications, with disagreements about both the selection of test problems and the choice of criteria for assessing computational results. This talk will briefly survey the current state of the art, trying along the way to highlight a few of the interesting open questions.

BIOGRAPHY

Margaret H. Wright is Silver Professor of Computer Science and Mathematics in the Courant Institute of Mathematical Sciences, New York University. She received her B.S. (Mathematics) and M.S. and Ph.D. (Computer Science) from Stanford University. Her research interests include optimization, scientific computing, and optimization in real-world applications. Prior to joining NYU, she worked at Bell Laboratories (Lucent Technologies), where she was named as a Bell Labs Fellow. She was elected to the National Academy of Engineering (1997), the American Academy of Arts and Sciences (2001), and the National Academy of Sciences (2005). During 1995-1996 she served as president of the Society for Industrial and Applied Mathematics (SIAM), and she has chaired advisory committees for several mathematical sciences institutes and government agencies. In 2000, she received an honorary doctorate in Mathematics from the University of Waterloo, and she was named as an honorary doctor of technology by the Swedish Royal Institute of Technology in 2008.


Dr. James Ostrowski, Post Doctoral Fellow

Title: Symmetry in Integer Programming

University of Waterloo, Ontario, Canada

Friday, January 29, 2010

3:00 p.m. - 4:00 p.m.

Mohler Lab Room 451

ABSTRACT

Symmetry has long been considered a curse in integer linear programming, and auxiliary formulations are often sought to reduce the amount of symmetry a formulation contains. The standard method for solving integer programs is branch-and-bound. In branch-and-bound, the set of feasible solutions is partitioned, forming more easily-solved subproblems. The presence of symmetry means that many of these subproblems are equivalent. In this talk, we will discuss how to recognize and exploit the symmetry when solving integer linear programs via branch-and-bound.

BIOGRAPHY

Jim Ostrowski is a postdoctoral fellow at the University of Waterloo. He received his Ph.D. in the Industrial and Systems Engineering department at Lehigh University in 2009 and M.A. degrees in both mathematics and statistics from Miami University, Ohio, in 2004. His research interests lie in discrete and nonlinear optimization, with recent application in the power industry.


Dr. Uday V. Shanbhag

Title: Nash Games Under Uncertainty: Characterization and Computation With Applications to Power Markets

Industrial and Enterprise Systems Engineering, The University of Illinois at Urbana-Champaign

Friday, January 15, 2010

2:30 p.m. - 3:30 p.m.

Mohler Lab Room 451

ABSTRACT

While early work by Dantzig and Beale made the first inroads into stochastic programming, game-theoretic extensions, namely stochastic Nash games, have been less studied. Networked extensions of such games, where strategy sets are coupled and objectives are nonsmooth, pose unique challenges. Motivated by risk-averse bidding problems in power markets, we consider the dual questions of characterizing equilibria and their computation via scalable distributed schemes. Characterization questions are addressed through variational techniques, whose application is complicated by the nonsmoothness arising from the use of risk measures. A novel projection-based cutting plane scheme is presented for purposes of computation, along with rate estimates and error bounds, and is shown to scale well with problem size. Time permitting; we will discuss some related work that relates the solvability of a sampled Nash game to that of the original stochastic Nash game.

BIOGRAPHY

Uday V. Shanbhag is an assistant professor in Industrial and Enterprise Systems Engineering at the University of Illinois at Urbana-Champaign. His interests lie in development of theory and algorithms for optimization and game-theoretic problems, particularly in uncertain and dynamic settings, with application interests in the design and operation of electricity markets. He has a Ph.D. from Stanford University's department of Management Science and Engineering in 2006 and also holds S.M. and B. Tech degrees from MIT and IIT, Mumbai, respectively. His research awards include the triennial A.W. Tucker Prize by the mathematical programming society (MPS) in 2006 and the Computational Optimization and Applications (COAP), best paper award (jointly with Walter Murray) in 2007. He was also selected as one of 11 finalists for the Microsoft New Faculty fellowship (2007) and his research has been supported by the National Science Foundation (NSF, the Department of Energy (DOE) and the US Army.