Training neural networks using Taguchi methods: overcoming interaction problems.
Viswanathan, Alagappan; MacLeod, Christopher; Maxwell, Grant; Kalidindi, Sashank
Taguchi Methods (and other orthogonal arrays) may be used to train small Artificial Neural Networks very quickly in a variety of tasks. These include, importantly, Control Systems. Previous experimental work has shown that they could be successfully used to train single layer networks with no difficulty. However, interaction between layers precluded the successful reliable training of multi-layered networks. This paper describes a number of successful strategies which may be used to overcome this problem and demonstrates the ability of such networks to learn non-linear mappings.
|Start Date||Sep 11, 2005|
|Publication Date||Sep 1, 2005|
|Publisher||Springer (part of Springer Nature)|
|Series Title||Lecture notes in computer science|
|Institution Citation||VISWANATHAN, A., MACLEOD, C., MAXWELL, G. and KALIDINDI, S. 2005. Training neural networks using Taguchi methods: overcoming interaction problems. In Duch, W., Kacprzyk, J., Oja, E. and Zadrozy, S (eds.) Artificial neural networks: formal models and their applications, part 2; proceedings of the 15th International on artificial neural networks (ICAAN 2005), 11-15 September 2005, Warsaw, Poland. Lecture notes in computer science, 3697. Berlin: Springer, pages 103-108. Available from: https://dx.doi.org/10.1007/11550907_17|
|Keywords||Taguchi methods; Networks; Nonlinear mappings; Control systems|
VISWANATHAN 2005 Training neural networks using Taguchi