论文信息 - Growing nonuniform feedforward networks for continuous mappings

Growing nonuniform feedforward networks for continuous mappings

Abstract In this paper we propose a greedy algorithm for growing nonuniform feedforward neural networks for approximating continuous valued mappings. The convergence of the algorithm is rigorously proved. Experimental studies show that the algorithm can grow good networks structures for even complex functions.

Sujoy Ghose | V. V. Vinod

[1] Pietro Burrascano,et al. A norm selection criterion for the generalized delta rule , 1991, IEEE Trans. Neural Networks.

[2] Timur Ash,et al. Dynamic node creation in backpropagation networks , 1989 .

[3] J. Nadal,et al. Learning in feedforward layered networks: the tiling algorithm , 1989 .

[4] Louis A. Pipes,et al. Applied Mathematics for Engineers and Physicists , 1959 .

[5] James L. McClelland,et al. Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations , 1986 .

[6] Terence D. Sanger,et al. A tree-structured adaptive network for function approximation in high-dimensional spaces , 1991, IEEE Trans. Neural Networks.

[7] Mark Alan Fanty. Learning in structured connectionist networks , 1988 .

[8] R. Palmer,et al. Introduction to the theory of neural computation , 1994, The advanced book program.