Advances in Mathematica l Programming and Optimization in the SAS System

Optimization continues to grow in importance as the tools become more reliable and handle larger, more difficult problems. As a result, new applications with more variables and more complicated models are being introduced. The SAS System has numerous optimization routines which handle the standard problems such as linear and nonlinear programming, integer programming, network flows, linear and nonlinear regression with all types of constraints, as well as problems with highly specialized structure, such as linear programs with embedded networks. These capabilities are exposed to the users in a variety of places such as in the procedures LP, NLP, NLMIXED, and IML, and in applications like neural networks and regression in Enterprise Miner software. This presentation gives an overview of the optimization capabilities in the SAS System, how they are changing with the availability of important new techniques, and where they will be in future releases.