论文信息 - A Study on Functional-Link Neural Units with Maximum Entropy Response

A Study on Functional-Link Neural Units with Maximum Entropy Response

In this paper a preliminary study on an unsupervised learning theory for functional-link neural units is presented. A set of learning equations for this neural topology based on an information-theoretic approach is given. Then, learning and approximation capabilities are investigated through computer simulations concerning density shaping.

Simone Fiori | Francesco Piazza

[1] Mark D. Plumbley. Efficient information transfer and anti-Hebbian neural networks , 1993, Neural Networks.

[2] Ralph Linsker,et al. Local Synaptic Learning Rules Suffice to Maximize Mutual Information in a Linear Network , 1992, Neural Computation.

[3] Yuhong Yang,et al. An Asymptotic Property of Model Selection Criteria , 1998, IEEE Trans. Inf. Theory.

[4] Vladimir Vapnik,et al. The Support Vector Method , 1997, ICANN.

[5] Yoh-Han Pao,et al. Adaptive pattern recognition and neural networks , 1989 .

[6] Yoram Baram,et al. Multidimensional density shaping by sigmoids , 1996, IEEE Trans. Neural Networks.

[7] Perry Moerland. Mixtures of Experts Estimate A Posteriori Probabilities , 1997, ICANN.

[8] George Havas,et al. A Family of Perfect Hashing Methods , 1996, Comput. J..

[9] Terrence J. Sejnowski,et al. An Information-Maximization Approach to Blind Separation and Blind Deconvolution , 1995, Neural Computation.

[10] Mohamad H. Hassoun,et al. Nonlinear Hebbian rule: a statistical interpretation , 1994, Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94).

[11] Andrzej Cichocki,et al. Stability Analysis of Learning Algorithms for Blind Source Separation , 1997, Neural Networks.