Subset selection using nonlinear optimization

A common problem in computer science is how to represent a large dataset in a smaller more compact form. This thesis describes a generalized framework for selecting canonical subsets of data points that are highly representative of the original larger dataset. The contributions of the work are formulation of the subset selection problem as an optimization problem, an analysis of the complexity of the problem, the development of approximation algorithms to compute canonical subsets, and a demonstration of the utility of the algorithms in several problem domains.