The Empirical Behavior of Sampling Methods for Stochastic Programming

Jeff Linderoth, Alexander Shapiro, Stephen Wright

This is a companion web site for the paper

J. T. Linderoth, A. Shapiro, and S. J. Wright, The Empirical Behavior of Sampling Methods for Stochastic Programming, Optimization Technical Report 02-01, Computer Science Department, University of Wisconsin-Madison, January, 2002. Revised October, 2002.

See latest PDF version of the paper. This site contains additional details on the numerical tests, data sets used in the tests, and pointers to related information.

Data

We used the SMPS input format for our problems, which were gathered from various sources in the literature, online test problem archives, and colleagues.

To obtain the data, download these tar files and run tar xvf (filename).tar to recover a directory named data-problemname containing files with suffices .cor (core file), .sto, .tim, and .sol. The first three contain the input data in SMPS format. The last contains a suggested starting point-usually one that is relatively close to the true solution.

Details of Experiments

Currently the paper contains a fair amount of detail about the experiments. Eventually we'll put additional details here.

Links

Here are some links to sites with computational stochastic programming data, in various states of repair.

Support

  • The Department of Energy, Office of Advanced Scientific Computing Research
  • The National Science Foundation Grant CDA-9726385.