Function choice, resiliency and growth in genetic programming

In this paper we examine how the choice of functions in a genetic program (GP) affects the rate of code growth and the development of resilient individuals. We find that functions or combination of functions that produce the most resilient individuals often, but not always, produce the most growth. Finally, the results confirm that there is a high correlation between the resiliency of parent individuals and the fitness of their offspring. Categories and Subject Descriptors I.2.2 [Automatic Programming]: Program Synthesis General Terms Algorithms Keywords Genetic programming, function choice, growth, resiliency

• Type: Conference
• Year: 2005
• Where: GECCO
• Authors: Sireesha Besetti, Terence Soule
• Keywords: Code Growth, GECCO 2005, Keywords Genetic Programming, Resilient Individuals