Enhancing the Performance of GP Using an Ancestry-Based Mate Selection Scheme

The performance of genetic programming relies mostly on population-contained variation. If the population diversity is low then there will be a greater chance of the algorithm being unable to find the global optimum. We present a new method of approximating the genetic similarity between two individuals using ancestry information. We define a new diversity-preserving selection scheme, based on standard tournament selection, which encourages genetically dissimilar individuals to undergo genetic operation. The new method is illustrated by assessing its performance in a well-known problem domain: algebraic symbolic regression. 1 Ancestry-Based Tournament Selection Genetic programming [1] in its most conventional form, suffers from a well-known problem: the search becomes localised around a point in the search space which is not the global optimum. This may be a result of not taking steps to preserve the genetic diversity of the population. A selection scheme in GP may be made to promot...

• Type: Conference
• Year: 2003
• Where: GECCO
• Authors: Rodney Fry, Andrew M. Tyrrell
• Keywords: GECCO 2003, Genetic Programming, Global Optimum, Selection Scheme