2014 (modified: 11 Nov 2022)ICML 2014Readers: Everyone
Abstract:Representing examples in a way that is compatible with the underlying classifier can greatly enhance the performance of a learning system. In this paper we investigate scalable techniques for induc...