Title :
An incremental learning method with relearning of recalled interfered patterns
Author :
Yamauchi, Koichiro ; Yamaguchi, Nobuhiko ; Ishi, N.
Author_Institution :
Dept. of Intell. & Comput. Sci., Nagoya Inst. of Technol., Japan
Abstract :
This paper presents a new incremental learning method for neural networks. If a neural network is trained to memorize novel patterns only by their presentation, the network will forget some patterns that have been already learnt. This problem is caused by the fact that the learning of novel patterns usually interfere in the internal representation corresponding to the old training patterns. In the new method, the network recalls the patterns that the novel patterns possibly interfere in, and then learns both novel and recalled patterns. In the computer simulation, we demonstrate this system in the learning of alphabetic characters
Keywords :
learning (artificial intelligence); neural nets; pattern recognition; alphabetic characters; incremental learning method; internal representation; recalled interfered patterns; relearning; Buffer storage; Computational complexity; Computer science; Computer simulation; Humans; Intelligent networks; Learning systems; Neural networks; Paper technology; Pattern recognition;
Conference_Titel :
Neural Networks for Signal Processing [1996] VI. Proceedings of the 1996 IEEE Signal Processing Society Workshop
Conference_Location :
Kyoto
Print_ISBN :
0-7803-3550-3
DOI :
10.1109/NNSP.1996.548354