Synthesis of Sigma-Pi Neural Networks by the Breeder Genetic Programming

Created by W.Langdon from gp-bibliography.bib Revision:1.3973

@InProceedings{Zhang-Muehlenbein-94-WCCI-EC,
  author =       "Byoung-Tak Zhang and Heinz M{\"u}hlenbein",
  title =        "Synthesis of Sigma-Pi Neural Networks by the Breeder
                 Genetic Programming",
  booktitle =    "Proceedings of IEEE International Conference on
                 Evolutionary Computation (ICEC-94), World Congress on
                 Computational Intelligence",
  publisher =    "IEEE Computer Society Press",
  address =      "Orlando, Florida, USA",
  month =        "27-29 " # jun,
  publisher_address = "New York, USA",
  year =         "1994",
  pages =        "318--323",
  keywords =     "genetic algorithms, genetic programming",
  URL =          "http://bi.snu.ac.kr/Publications/Conferences/International/ICEC94.pdf",
  abstract =     "Genetic programming has been successfully applied to
                 evolve computer programs for solving a variety of
                 interesting problems. In the previous work we
                 introduced the breeder genetic programming (BGP) method
                 that has Occam's razor in its fitness measure to evolve
                 minimal size multilayer perceptrons. In this paper we
                 apply the method to synthesis of sigma-pi neural
                 networks. Unlike perceptron architectures, sigma-pi
                 networks use product units as well as summation units
                 to build higher-order terms. The effectiveness of the
                 method is demonstrated on benchmark problems.
                 Simulation results on noisy data suggest that BGP not
                 only improves the generalization performance, it can
                 also accelerate the convergence speed.",
  notes =        "Tests GP/Sigma-pi/NN on parity problems. On clean data
                 was able to produce S/P Neural Networks with high
                 performance >98% correct. Also ~90% on noisy
                 data.

                 Fitness function sums NN error and NN size/complexity
                 penalty terms. Shows size/complexity penalty beneficial
                 in that better NN are produced and the GP is twice as
                 fast.",
}

Genetic Programming entries for Byoung-Tak Zhang Heinz Muhlenbein

Citations