A dynamic mutation genetic algorithm
Tzung‐Pei Hong, Hong-Shung Wang · 2002
Conventional genetic algorithms use only one mutation operator to generate the next generation. Determining which mutation operator should be used is quite difficult and is usually done by trial-and-error. In this paper, a new genetic algorithm, the dynamic mutation genetic algorithm (DMGA), is proposed to solve the problem. The dynamic mutation genetic algorithm uses more than one mutation operator to generate the next generation. The mutation ratio of each operator changes along with the evaluation results from the respective offspring in the next generation. We thus expect that really good operators will have an increased effect on the genetic process.