Improving GP classifier generalization using a cluster separation metric

Abstract

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.

Improving generalization of evolved programs through automatic simplification

“…Various work has been done to reduce the problems of overtting in GP [4,[8][9][10]33]. In other work, an ensemble of symbolic regression datasets were created using bootstrapping [1].…”

“…While the challenges of learning solutions that generalize have long been recognized and studied, both in GP (e.g. [4,8,10,32,33]) and other machine learning algorithms (e.g. [5]), the challenges in this area have not yet been fully met, and signicant problems remain open.…”