An approach to control aging rate of neural networks under adaptation to gradually changing context

Author

Tanprasert, Thitipong ; Kripruksawan, Thosaporn

Author_Institution

Dept. of Comput. Sci., Assumption Univ. of Thailand, Bangkok, Thailand

Volume

1

fYear

2002

fDate

18-22 Nov. 2002

Firstpage

174

Abstract

The paper presents a decayed prior sampling algorithm for integrating the existing knowledge of a supervised learning neural networks with the new training data. The algorithm allows the existing knowledge to age out in slow rate as a neural network is gradually retrained with consecutive sets of new samples, resembling the change of application locality under a consistent environment. The experiments are performed on 2-dimensional partitions problem and the results convincingly confirm the effectiveness of the technique.

Keywords

adaptive systems; learning (artificial intelligence); neural nets; sampling methods; 2-dimensional partitions problem; application locality; consecutive sets; consistent environment; decayed prior sampling algorithm; neural network aging rate; neural network retraining; supervised learning neural networks; training data; Aging; Computer science; Contracts; Network synthesis; Neural networks; Neurons; Sampling methods; Speech recognition; Supervised learning; Training data;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Information Processing, 2002. ICONIP '02. Proceedings of the 9th International Conference on

Print_ISBN

981-04-7524-1

Type

conf

DOI

10.1109/ICONIP.2002.1202154

Filename

1202154