Optimal learning in artificial neural networks: A review of theoretical results

被引:34
作者
Bianchini, M [1 ]
Gori, M [1 ]
机构
[1] UNIV FLORENCE,DIPARTIMENTO SISTEMI & INFORMAT,I-50139 FLORENCE,ITALY
关键词
learning algorithms; optimal learning; connectionist models;
D O I
10.1016/0925-2312(95)00032-1
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The effectiveness of connectionist models in emulating intelligent behaviour and solving significant practical problems is strictly related to the capability of the learning algorithms to find optimal or near-optimal solutions and to generalize to new examples, This paper reviews some theoretical contributions to optimal learning in the attempt to provide a unified view and give the state of the art in the field. The focus of the review is on the problem of local minima in the cost function that is likely to affect more or less any learning algorithm. Starting from this analysis, we briefly review proposals for discovering optimal solutions and suggest conditions for designing architectures tailored to a given task.
引用
收藏
页码:313 / 346
页数:34
相关论文
共 84 条
[71]   LEARNING COMPLEX, EXTENDED SEQUENCES USING THE PRINCIPLE OF HISTORY COMPRESSION [J].
SCHMIDHUBER, J .
NEURAL COMPUTATION, 1992, 4 (02) :234-242
[72]  
Shynk J J, 1990, IEEE Trans Neural Netw, V1, P268, DOI 10.1109/72.80252
[73]  
Solla S. A., 1988, Complex Systems, V2, P625
[74]  
SONTAG E, 1989, INT JOINT C NEUR NET, V1, P639
[75]  
Sontag E. D., 1989, Complex Systems, V3, P91
[76]  
TOM, 1987, LECT NOTES COMPUTER
[77]   PHONEME RECOGNITION USING TIME-DELAY NEURAL NETWORKS [J].
WAIBEL, A ;
HANAZAWA, T ;
HINTON, G ;
SHIKANO, K ;
LANG, KJ .
IEEE TRANSACTIONS ON ACOUSTICS SPEECH AND SIGNAL PROCESSING, 1989, 37 (03) :328-339
[78]  
Werbos P. J., 1974, Ph.D. Thesis
[79]   30 YEARS OF ADAPTIVE NEURAL NETWORKS - PERCEPTRON, MADALINE, AND BACKPROPAGATION [J].
WIDROW, B ;
LEHR, MA .
PROCEEDINGS OF THE IEEE, 1990, 78 (09) :1415-1442
[80]  
Widrow B., 1960, 1960 IRE WESCON CONV, V4, P96