论文信息 - Stability of Equilibrium Points and Storage Capacity of Hopfield Neural Networks with Higher Order Nonlinearity

Stability of Equilibrium Points and Storage Capacity of Hopfield Neural Networks with Higher Order Nonlinearity

In this paper, we consider the storage capacity and stability of the so-called Hopfield neural networks with higher order nonlinearity. There are different ways to introduce higher order nonlinearity to the network; however we have considered one which does not have a huge computational cost. It is shown that, this modification of the Hopfield model significantly improves the storage capacity. We also classify the model via a stability measure, and study the effect of training the network with biased patterns on the stability.

Mohammad Bagher Menhaj | Mohammad Reza Rajati

[1] Heng-Ming Tai,et al. Neural networks with higher-order nonlinearity , 1988 .

[2] W. Krauth,et al. Learning algorithms with optimal stability in neural networks , 1987 .

[3] C. L. Giles,et al. High order correlation model for associative memory , 1987 .

[4] Richard O. Duda,et al. Pattern classification and scene analysis , 1974, A Wiley-Interscience publication.

[5] Demetri Psaltis,et al. Higher order associative memories and their optical implementations , 1988, Neural Networks.

[6] Thomas B. Kepler,et al. Universality in the space of interactions for network models , 1989 .

[7] Santosh S. Venkatesh,et al. The capacity of the Hopfield associative memory , 1987, IEEE Trans. Inf. Theory.

[8] Baldi,et al. Number of stable points for spin-glasses and neural networks of higher orders. , 1987, Physical review letters.

[9] Mohammad Bagher Menhaj,et al. On Classification of Some Hopfield-Type Learning Rules via Stability Measures , 2006 .

[10] David Burshtein. Long-term attraction in higher order neural networks , 1998, IEEE Trans. Neural Networks.

[11] C. L. Giles,et al. Machine learning using higher order correlation networks , 1986 .