Gaussian mutation and self-adaption for numeric genetic algorithms

By considering the function variables rather than the binary-bits as genes, new mutation operators can be devised for GAs used to optimise numeric functions. We implement Gaussian mutation operators for genetic algorithms used to optimise numeric functions and show it is superior to bit-flip mutation for most of the test functions. Gaussian mutation is a fundamental operator of both evolutionary strategies (ES) and evolutionary programming (EP). We also implement self-adaptive Gaussian mutation (also used in evolutionary strategies and evolutionary programming) which allows the GA to vary the mutation strength during the run, this gives further improvement on some of the functions. The performance of our GA using a simple implementation of self-adaptive Gaussian mutation is now comparable to ESs. This shows the importance of mutation and the importance of using appropriate mutation operators