Learning recursive functions from noisy examples using generic genetic programming

摘要: One of the most important and challenging areas research in evolutionary algorithms is investigation ways to successfully apply larger more complicated problems. In this paper, we GGP (Generic Genetic Programming) evolve general recursive functions for even-n-parity problem from noisy training examples. very flexible programs various programming languages can be acquired. Moreover, it powerful enough handle context-sensitive information domain-dependent knowledge. A number experiments have been performed determine impact noise examples on speed learning.