Population Learning Metaheuristic for Neural Network Training
Population based methods handle a population of individuals that evolves with the help of information exchange and self-improvement procedures. In this paper an application of a new metaheuristic called population learning algorithm (PLA) to ANN is investigated. The paper proposes several implementations of the PLA to training feed-forward artificial neural networks. The approach is validated by means of computational experiment in which PLA algorithm is used to train ANN solving a variety of benchmarking problems. Results of the experiment prove that PLA can be considered as a useful and effective tool for training ANN.
KeywordsInitial Population Neural Network Training Wisconsin Breast Cancer Neural Network Training Algorithm Global Minimization Method
Unable to display preview. Download preview PDF.
- 1.Duch, W., Korczak, J. (1998): Optimization and Global Minimization Methods Suitable for Neural Network, Neural Computing Surveys 2, http://www.icsi.berkeley.edu/jagopta/CNS
- 2.Fahlman, S.E., Lebiere, C. (1990): The Cascade-Corelation Learning Architecture, in Touretzky (Ed.) Advances in Neural Information Processing Systems 2, Morgan KaufmannGoogle Scholar
- 3.Hertz, J., Krogh, A., Palmer, R.G. (1995): Introduction to the Theory of Neural Computation, WNT, Warsaw (In Polish)Google Scholar
- 4.Jgdrzejowicz, P. (1999): Social Learning Algorithm as a Tool for Solving Some Difficult Scheduling Problems, Foundation of Computing and Decision Sciences 24, 51–66Google Scholar
- 5.Kevin, J., Witbrock, L., Witbrock, M.J. (1998): Learning to Tell Two Spirals Adapt, in Proceedings of the Connectionist Models Summer School, Morgan KaufmannGoogle Scholar
- 6.Mangasarian, O.L., Wolberg, W.H. (1990): Cancer Diagnosis via Linear Programming, SIAM News, vol. 23, Number 5, 1–18Google Scholar
- 8.Prechelt, L. (1994): Proben 1 — A Set of Benchmark and Benchmarking Rules for Neural Network Training Algorithm, Technical Report 21/94, Fakultt fr Informatik, Universitt Karlsruhe, D-76128 Karlsruhe, Germany, Anonymous /pub/papers/techraports/1994/199421.ps.Z. on ftp.ira.uka.deGoogle Scholar
- 9.Wah, B.W., Minglun Qian (2000): Constrained Formulations for Neural Network Training and Their Applications to Solve the Two-Spiral Problem, Proc. of Fifth Int’l Conference on Computer Science and Informatics, ICIS, 1, 598–601Google Scholar
- 10.Yi Shang, Wah, B.W. (1996): A Global Optimization Method for Neural Network Training, Conference of Neural Networks, IEEE Computer 29, 45–54Google Scholar