Recent advances in computing power have dramatically stimulated the development of methods in statistics and system optimization. Much of the progress is due to an increased ability to optimize general multiparameter functions--partly as a result of increased computing power in general, but more because of algorithmic advances in optimization techniques. So far there has been only sporadic informal contact between those who develop new methodology and those who develop optimization methods. This project will support a formal collaboration of reseatchers in the departments of Computer Science, Electrical Engineering, Operations Research, and Statistics. Objectives include the development and implementation of algorithms in the areas of projection pursuit, global optimization, multivariate smoothing, and decentralized controller design. Associated with the methodological development will be research into theoretical and inferential properties, combined with issues of interpretation and graphic display.
