Mutation in Compressed Encoding in Estimation of Distribution Algorithm
Orawan Watchanupaporn, Worasait Suwannik, Prabhas Chongstitvatana · 2012
Estimation of Distribution Algorithm (EDA) is a new kind of evolutionary algorithm. However, it does not use evolutionary operators such as crossover and mutation. in this paper, we investigate how mutation has an effect on the performance of EDA, more specifically, compact genetic algorithm (cGA) and LZWcGA, the latter uses compressed encoding. the result shows that cGA performs poorly with mutation while LZWcGA's performance is improved by mutation. We also present an analysis of mutation in both algorithms.