Efficient learning in adaptive processing of data structures

被引：16

作者：

Cho, SY ^{[1
]}

Chi, ZR ^{[1
]}

Wang, ZY ^{[1
]}

Siu, WC ^{[1
]}

机构：

[1] Hong Kong Polytech Univ, Dept Elect & Informat Engn, Ctr Multimedia Signal Proc, Kowloon, Hong Kong, Peoples R China

来源：

NEURAL PROCESSING LETTERS | 2003年 / 17卷 / 02期

关键词：

adaptive processing of data structures; backpropagation through structures; long-term dependency problem;

D O I：

10.1023/A:1023680629977

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Many researchers have explored the use of neural network models for the adaptive processing of data structures. The learning formulation for one of the models is known as the Backpropagation Through Structure (BPTS) algorithm. The main limitations of the BPTS algorithm are attributed to the problems of slow convergence speed and long-term dependency. In this Letter, a novel heuristic algorithm is proposed. The idea of this algorithm is to optimize the free parameters of the node representation in data structure by using a hybrid type of learning algorithm. Encouraging results achieved demonstrate that this proposed algorithm outperforms the BPTS algorithm.

引用

页码：175 / 190

页数：16

共 13 条

[1]

AARTS EH, 1999, SIMULATED ANNEALING

[2] TRAINING NEURAL NETS WITH THE REACTIVE TABU SEARCH [J].

BATTITI, R ;

TECCHIOLLI, G .

IEEE TRANSACTIONS ON NEURAL NETWORKS, 1995, 6 (05) :1185-1200

[3] LEARNING LONG-TERM DEPENDENCIES WITH GRADIENT DESCENT IS DIFFICULT [J].

BENGIO, Y ;

SIMARD, P ;

FRASCONI, P .

IEEE TRANSACTIONS ON NEURAL NETWORKS, 1994, 5 (02) :157-166

[4] Training multilayer neural networks using fast global learning algorithm - least-squares and penalized optimization methods [J].

Cho, SY ;

Chow, TWS .

NEUROCOMPUTING, 1999, 25 (1-3) :115-131

[5]

DODD N, 1990, INT JOINT C NEUR NET

[6] A general framework for adaptive processing of data structures [J].

Frasconi, P ;

Gori, M ;

Sperduti, A .

IEEE TRANSACTIONS ON NEURAL NETWORKS, 1998, 9 (05) :768-786

[7]

Goller C, 1996, IEEE IJCNN, P347, DOI 10.1109/ICNN.1996.548916

[8] ON THE PROBLEM OF LOCAL MINIMA IN BACKPROPAGATION [J].

GORI, M ;

TESI, A .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1992, 14 (01) :76-86

[9]

Rumelhart DE, 1986, PARALLEL DISTRIBUTED, V1, DOI DOI 10.7551/MITPRESS/5236.001.0001

[10] Global optimization for neural network training [J].

Shang, Y ;

Wah, BW .

COMPUTER, 1996, 29 (03) :45-+

← 1 2 →