Elegance: Genetic Algorithms in Neural Reinforcement Control
Pieter H.M. Spronck · 2001
Genetic algorithms are search algorithms based on the principles of natural selection and natural genetics, which are particularly suited for optimisation problems. Genetic algorithms work with a set of potential solutions to a problem, which all have been awarded a fitness rating, which is a measure of their relative success in solving the problem at hand. By using genetic operators and the fitness ratings, new potential solutions are created from the existing ones, in such a way that chances are that the features which make the better ones work, are passed on to their offspring.