论文信息 - Improving ANN generalization using a priori knowledge to pre-structure ANNs

Improving ANN generalization using a priori knowledge to pre-structure ANNs

This is a continuation of work reported by Lendaris at el. (1994) whose objective has been to develop a method that uses certain a priori information about a problem domain to pre-structure artificial neural networks (ANNs) into modules before training. The method is based on a general systems theory methodology, based on information-theoretic ideas, that generates structural information of the problem domain by analyzing I/O pairs from that domain. The notion of performance subset of an ANN structure is described. Extensive experiments on 5-input/1-output and 7-input/1-output Boolean mappings show that significantly improved generalization follows from successful pre-structuring. As the previous work already showed, such pre-structuring also yields improved training speed.

George G. Lendaris | A. Rest | T. R. Misley

[1] Klaus Krippendorff,et al. Information Theory: Structural Models for Qualitative Data. , 1988 .

[2] George G. Lendaris,et al. On Matching ANN Structure to Problem Domain Structure , 1993 .

[3] George G. Lendaris,et al. Structure and constraint in discrete adaptive networks , 1965 .

[4] Tsu-Chang Lee,et al. Structure level adaptation for artificial neural networks , 1991 .

[5] G. Lendaris,et al. On Prestructuring ANNs Using A Priori Knowledge , 1994 .

[6] Roger C. Conant,et al. EXTENDED DEPENDENCY ANALYSIS OF LARGE SYSTEMS , 1988 .

[7] George G. Lendaris,et al. Use of a structured problem domain to explore development of modularized neural networks , 1992, [Proceedings 1992] IJCNN International Joint Conference on Neural Networks.

[8] G. Lendaris,et al. Using A Priori Knowledge to Prestructure ANNs , 1994 .

[9] Dušan Petrovački,et al. Evolutional development of a multilevel neural network , 1993, Neural Networks.

[10] Russell Reed,et al. Pruning algorithms-a survey , 1993, IEEE Trans. Neural Networks.

[11] George G. Lendaris. A proposal for indicating quality of generalization when evaluating ANNs , 1990, 1990 IJCNN International Joint Conference on Neural Networks.

[12] David E. Rumelhart,et al. BACK-PROPAGATION, WEIGHT-ELIMINATION AND TIME SERIES PREDICTION , 1991 .

[13] Anders Krogh,et al. Introduction to the theory of neural computation , 1994, The advanced book program.