HIERARCHICAL NEURAL NETWORKS

被引:65
作者
MAVROVOUNIOTIS, ML [1 ]
CHANG, S [1 ]
机构
[1] UNIV MARYLAND,SYST RES CTR,COLLEGE PK,MD 20742
基金
美国国家科学基金会;
关键词
D O I
10.1016/0098-1354(92)80053-C
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
With the common three-layer neural network architectures, networks lack internal structure; as a consequence, it is very difficult to discern characteristics of the knowledge acquired by a network in order to evaluate its reliability and applicability. An alternative neural-network architecture is presented, based on a hierarchical organization. Hierarchical networks consist of a number of loosely-coupled subnets, arranged in layers. Each subnet is intended to capture specific aspects of the input data. A subnet models a particular subset of the input variables, but the exact patterns and relationships among variables are determined by training the network as a whole. However, the hierarchy of subnets gives the network hints to look for patterns in the most promising directions. Their modular organization makes hierarchical neural networks easier to analyze, because one can focus on the analysis of one subnet at a time, rather than attempt to decipher the whole network at once.
引用
收藏
页码:347 / 369
页数:23
相关论文
共 16 条
[1]  
BHAT N, 1990, COMPUT CHEM ENG, V14, P561
[2]   Product Units: A Computationally Powerful and Biologically Plausible Extension to Backpropagation Networks [J].
Durbin, Richard ;
Rumelhart, David E. .
NEURAL COMPUTATION, 1989, 1 (01) :133-142
[3]  
FOSTER B, 1990, AICHE ANN M CHICAGO
[4]   ANALYSIS OF HIDDEN UNITS IN A LAYERED NETWORK TRAINED TO CLASSIFY SONAR TARGETS [J].
GORMAN, RP ;
SEJNOWSKI, TJ .
NEURAL NETWORKS, 1988, 1 (01) :75-89
[5]   ARTIFICIAL NEURAL NETWORK MODELS OF KNOWLEDGE REPRESENTATION IN CHEMICAL-ENGINEERING [J].
HOSKINS, JC ;
HIMMELBLAU, DM .
COMPUTERS & CHEMICAL ENGINEERING, 1988, 12 (9-10) :881-890
[6]   IMPROVEMENT OF THE BACKPROPAGATION ALGORITHM FOR TRAINING NEURAL NETWORKS [J].
LEONARD, J ;
KRAMER, MA .
COMPUTERS & CHEMICAL ENGINEERING, 1990, 14 (03) :337-341
[7]  
Lippman R. P., 1987, IEEE ASSP MAGAZI APR, P4
[8]  
McClelland J. L., 1989, HDB MODELS PROGRAMS
[9]  
MOODY J, 1988, 1988 P CONN MOD SUMM
[10]  
Sejnowski T. J., 1987, Complex Systems, V1, P145