We will present in this paper a demonstration that genetic search can be studied as an entropic phenomenon. We look in depth at the performance differences between the traditional Genetic Algorithm (GA) and the Enhanced Genetic Algorithm (EGA) based on measures of statistical entropy and information. Specifically, for the purpose of theoretical research, we use the Royal Road functions to demonstrate what we expect measures of entropy and information to provide and then we discuss the application of our entropy and information hypothesis to the Traveling Salesman Problem.
展开▼