论文信息 - Stochastic Hebbian learning with binary synapses

Stochastic Hebbian learning with binary synapses

This paper explores a variant of Hebbian learning in which binary synapses are updated stochastically rather than deterministically. In this variant, a single potentiation or depression event is implemented by setting a synapse weight respectively to "one" or "zero" with a finite probability, if it is not this value already. This learning rule is compared to the conventional Hebbian rule where a continuously valued synapse moves a fraction towards 1.0 or 0.0. It is shown that given a set of input-output pattern pairs, the expected value of a particular synapse is the same for both learning rules. Also, as the network size and the input activity levels increase, the signal to noise ratio of the dendritic sums approaches infinity. These stochastic binary synapses are presented as a viable mechanism for the VLSI implementation of Hebbian-based neural networks.

G. L. Barrows

[1] D. O. Hebb,et al. The organization of behavior , 1988 .

[2] Kyusun Choi Kyusun Choi,et al. Fully Parallel Digital Implementation Of Connectionist Associative Memory , 1992, 1992 Proceedings of the IEEE Custom Integrated Circuits Conference.

[3] H. C. LONGUET-HIGGINS,et al. Non-Holographic Associative Memory , 1969, Nature.

[4] Martin A. Brooke,et al. Low control voltage programming of floating gate MOSFETs and applications , 1994 .

[5] Carver A. Mead,et al. A single-transistor silicon synapse , 1996 .

[6] Tadashi Shibata,et al. A neuron-MOS neural network using self-learning-compatible synapse circuits , 1995, IEEE J. Solid State Circuits.

[7] Michel Declercq,et al. Implementation of a learning Kohonen neuron based on a new multilevel storage technique , 1991 .

[8] H. Shinohara,et al. A refreshable analog VLSI neural network chip with 400 neurons and 40 k synapses , 1992, 1992 IEEE International Solid-State Circuits Conference Digest of Technical Papers.

[9] Manfred Glesner,et al. A Cascadable VLSI Architecture for the Realization of Large Binary Associative Networks , 1991 .