Improving GP classifier generalization using a cluster separation metric

Genetic Programming offers freedom in the definition of the cost function that is unparalleled among supervised learning algorithms. However, this freedom goes largely unexploited in previous work. Here, we revisit the design of fitness functions for genetic programming by explicitly considering the contribution of the wrapper and cost function. Within the context of supervised learning, as applied to classification problems, a clustering methodology is introduced using cost functions which encourage maximization of separation between in and out of class exemplars. Through a series of empirical investigations of the nature of these functions, we demonstrate that classifier performance is much more dependable than previously the case under the genetic programming paradigm. Categories and Subject Descriptors I.2.2 [Artificial Intelligence]: Automatic Programming General Terms Algorithms, Experimentation, Performance Keywords genetic programming, clustering, classification, evaluation

• Type: Conference
• Year: 2006
• Where: GECCO
• Authors: Ashley George, Malcolm I. Heywood
• Keywords: Cost Functions, GECCO 2006, Genetic Programming, Optimization, Supervised Learning