Matthew Tanner
Department of Industrial and Systems Engineering
College Station, Texas 77840
office: 304 Zachry Hall
email: mtanner at tamu dot edu
Curriculum Vitae
Adviser: Dr. Lewis Ntaimo website
 

Education:4th year Ph.D. student in the Industrial and Systems Engineering Department at Texas A&M University
B.S.E. from the Operations Research and Financial Engineering Department at Princeton University in 2004   

· Research Interests and Applications
1.  I am currently researching new algorithms and MIP techniques for the optimization of joint chance-constrained stochastic programs with discretely distributed random variables.  The specific class of problems that I am studying is problems with randomness in the constraint matrix as well as the righthand side.  Until now, this class of problems has been considered particularly intractable because the only general solution method is a "big M" MIP formulation with an extremely weak relaxation.  The goal of my research is to derive new MIP methods for solving this problem including new classes of cutting planes, a stronger convex relaxation, and good metaheuristics for finding feasible solutions.  I also focus on implementing and testing these algorithms.

2.  I am also interesting in developing and implementing new decomposition and cutting plane algorithms for two-stage stochastic mixed-integer programs.  I have been studying the application of disjunctive programming techniques to two-stage stochastic MIPs.  The advantage of disjunctive programming is that cutting planes developed for one scenario subproblem can be quickly translated and made valid for all the scenario subproblems.  This enables strong disjunctive cuts to be derived relatively cheaply computationally speaking.

3.  One application that I am particularly interested in is finding optimal vaccinations strategies under uncertainty.  The problem of allocating scarce vaccine supplies effectively is important for preventing disease epidemics.  Most current research has focused on methods for finding the optimal allocation without taking into account the extreme uncertainty in the parameter values of disease epidemic models.  I am applying stochastic programming techniques in an attempt to derive more robust vaccination strategies than can be found at the moment 

Papers
1.  Ntaimo, L. and M. Tanner, "Computations with Disjunctive Cuts for Two-Stage Stochastic Mixed 0-1 Integer Programs," J. of Global Optimization (2008) 41: 365-384.  PDF
2.  Tanner, M., L. Sattenspiel, and L. Ntaimo, "Finding Optimal Vaccination Policies Under Parameter Uncertainty Using Stochastic Programming," Math. Biosciences doi:10.1016/j.mbs.2008.07.006. PDF
3.  Tanner, M. and E. Beier, "A General Heuristic Method for Joint Chance-Constrained Stochastic Programs with Discretely Distributed Parameters," under revision. PDF
4.  Tanner, M., L. Ntaimo, "IIS Branch-and-Cut for Joint Chance-Constrained Programs with Random Technology Matrices" under review. PDF 

· Conference Presentations 
1.  "IIS Cuts for Chance Constrained Stochastic Programs: Implementation and Computational Results," INFORMS SW Regional Conference, College Station, TX, April 19, 2008.
2.  "A Scenario Based Branch-and-Cut Algorithm for Stochastic Programming with Probabilistic Constraints," INFORMS Annual Meeting, Seattle, WA, Nov. 4-7, 2007.
3.  “A Computational Study of Lift-and-Project Cuts for Stochastic Mixed 0-1 Programs,” INFORMS International Meeting, Puerto Rico, July 8-11, 2007.
4.  “A Comparative Study of Disjunctive Cutting Planes for Two-Stage Stochastic Mixed-Binary Programs,” INFORMS Annual Meeting, Pittsburgh, PA, Nov. 5-8, 2006.
5.  “Assessing the Cost of Parameter Uncertainty when Finding Optimal Vaccination Strategies,” Applied Optimization and Metaheuristic Innovations, Yalta, Ukraine, July 19-21, 2006. 

· Teaching
1.  ISEN 302, Engineering Economy, Fall 2008
2.  ISEN 303, Engineering Economy, Summer 2008
3.  ISEN 302, Engineering Economy, Spring 2008
4.  Lab sections for INEN 314, Statistical Quality Control from Fall 2005 - Spring 2007 

Links
1. Community of Stochastic Programming COSP