Published June 1, 2007 | Version v1
Journal article Open

Búsqueda de la estructura óptima de redes neurales con Algoritmos Genéticos y Simulated Annealing. Verificación con el benchmark PROBEN1

Authors/Creators

Description

Este artı́culo describe el uso de algoritmos genéticos (AG) y simulated annealing (SA) en la búsqueda
de configuraciones óptimas de redes neurales artificiales, dentro de una arquitectura software, TSAGANN.
El estudio comparativo ha sido realizado con benchmarks consolidados y es ilustrado en detalle. El análisis
estadı́stico de los resultados indica que SA es tan eficiente como AG para este tipo de problemas, permitiendo
incluso realizar exploraciones en el espacio del problema con un menor número de evaluaciones de las usadas
por el AG para obtener resultados comparables.

Files

tsagann-paper_es.pdf

Files (339.2 kB)

Name Size Download all
md5:df127080f249e0f3d81770e55d188609
339.2 kB Preview Download

Additional details

References

  • Yas Abbas Alsultanny and Musbah M. Aqel. Pattern recognition using multilayer neural- genetic algorithm. Neurocomputing, 51:237– 247, 2003.
  • Karthik Balakrishnan and Vasant Honavar. Properties of genetic representations of neu- ral architectures. In Proceedings of the World Congress on Neural Networks, pages 807– 813, 1995.
  • David Beasley, David R. Bull, and Ralph R. Martin. An overview of genetic algorithms: Part 1 fundamentals. University Computing, 15(2):58–69, 1993.
  • David Beasley, David R. Bull, and Ralph R. Martin. An overview of genetic algorithms: Part 2 research topics. University Compu- ting, 15(4):170–181, 1993.
  • Christopher M. Bishop. Neural Networks for Pattern Recognition. Oxford University Press, 1995.
  • Erick Cantú-Paz. Pruning neural networks with distribution estimation algorithms. In Erick Cantú-Paz et al., editor, Genetic and Evolutionary Computation Conference, GECCO 2003, pages 790–800, 2003.
  • Erick Cantú-Paz and Chandrika Kamath. Evolving neural networks for the classifica- tion of galaxies. In Proceedings of the Ge- netic and Evolutionary Computation Confe- rence, GECCO 2002, pages 1019–1026, San Francisco, 2002. Morgan Kaufmann Publis- hers.
  • Erick Cantú-Paz and Chandrika Kamath. Evolving neural networks to identify bent-double galaxies in the FIRST survey. Neural Networks, 16(3-4):507–517, 2003.
  • Erick Cantú-Paz and Chandrika Kamath. An empirical comparison of combinations of evolutionary algorithms and neural networks for classification problems. In IEEE Tran- sactions on Systems, Man, and Cybernetics- Part B: Cybernetics, pages 915–927, 2005.
  • Stephanie Forrest. Genetic algorithms: Prin- ciples of natural selection applied to compu- tation. Science, 261:872–878, August 1993.
  • Stephanie Forrest and Melanie Mitchell. What makes a problem hard for a genetic al- gorithm? Some anomalous results and their explanation. Machine Learning, 13:285–319, 1993.
  • John H. Holland. Adaptation in Natural and Artificial Systems. MIT Press, first edition, 1992.
  • John H. Holland. Algoritmi genetici. Scienze, 289:50–57, September 1992.
  • Terry Jones and Stephanie Forrest. Fitness distance correlation as a measure of problem difficulty for genetic algorithms. In Larry Eshelman, editor, Proceedings of the Sixth International Conference on Genetic Algo- rithms, pages 184–192, San Francisco, CA, 1995.
  • Terry Jones and Stephanie Forrest. Genetic algorithms and heuristic search. In Procee- dings of International Joint Conference on Artificial Intelligence, 1995.
  • S. Kirkpatrick, C. D. Gelatt Jr., and M. P. Vecchi. Optimization by simulated annea- ling. Science, 220:671–680, 1983.
  • Ben Kröse and Patrick van der Smagt. An introduction to neural networks. University of Amsterdam, 1996.
  • Helmut A. Mayer, Reinhold Huber, and Ro- land Schwaiger. Lean artificial neural ne- tworks - regularization hLutz Prechelt. Proben1 - a set of neural ne- twork benchmark problems and benchmar- king rules. Technical Report 21/94, Fakültat für Informatik, Universität Karlsruhe, 76128 Karlsruhe, Germany, September 1994.elps evolution. In Proceedings of the 2NWGA, Second Nordic Workshop on Genetic Algorithms, 1996.
  • William H. Press, Saul A. Teukolsky, Wi- lliam T. Vetterling, and Brian P. Flannery. Numerical Recipes in C, The Art of Scienti- fic Computing. Cambridge University Press, second edition, 1995.
  • David E. Rumelhart, Geoffrey E. Hinton, and Ronald J. Williams. Learning internal representations by error propagation, volu- me Parallel Distributed Processing: Explo- rations in the Microstructures of Cognition, Vol. 1. 1986. pp. 318-362.
  • David E. Rumelhart, Geoffrey E. Hinton, and Ronald J. Williams. Learning represen- tations by back-propagation errors. Nature, 323:533–536, 1986.
  • Wolfram Schiffmann. Encoding feedfo- rward networks for topology optimization by simulated evolution. In Proceedings of the Fourth International Conference on Knowledge-Based Intelligent Systems and Allied Technologies, volume 1, pages 361– 364, 2000.
  • Randall S. Sexton, Robert E. Dorsey, and John D. Johnson. Towards global optimiza- tion of neural networks: a comparison of the genetic algorithm and backpropagation. De- cision Support Systems, 22(2):171–185, 1998.
  • William M. Spears. Adapting crossover in evolutionary algorithms. In Proceedings of the Evolutionary Programming Conference, pages 367–384, 1995.
  • Diomidis D. Spinellis and Chrissoleon T. Pa- padopoulos. Production line buffer alloca- tion: Genetic Algorithms versus Simulated Annealing. In Second International Aegean Conference on the Analysis and Modelling of Manufacturing Systems, pages 89–101, Ti- nos, Greece, 1999. University of the Aegean, Department of Business Administration.
  • Xin Yao. A review of evolutionary artificial neural networks. International Journal of In- telligent Systems, 8:539–567, 1993.
  • Xing Yao. Evolving artificial neural ne- tworks. In Proceedings of the IEEE, volu- me 87, pages 1423–1447, September 1999.
  • Xing Yao and Yong Liu. Towards des- igning artificial neural networks by evolu- tion. Applied Mathematics and Compu- tation, 91(1):83–90, 1998.
  • Ping Zhang, Brijesh Verma, and Kuldeep Kumar. Neural vs. statistical classifier in conjunction with genetic algorithm based feature selection. Pattern Recognition Let- ters, 26(7):909–919, 2005.