Associative Memories Using Multilayer Perceptrons with 3-Valued Weights and Sparsely Interconnected Neural Networks
Associative memories composed of sparsely interconnected neural networks (SINNs) are suitable for hardware implementation. However, the sparsely interconnected structure also gives rise to a decrease in the capability of SINNs for associative memories. Although this problem can be solved by increasing the number of interconnections, the hardware cost goes up rapidly. Therefore, we propose associative memories using multilayer perceptions (MLPs) with 3-valued weights and SINNs. This is because such MLPs can be realized at a lower cost than increasing interconnections in SINNs and can give each neuron in SINNs the global information of an input pattern to improve the storage capacity. Finatty, it is confirmed by simulations that our proposed associative memories have good performance.
Takeshi KAMIO Mititada MORISUE
Department of Information Machines and Interfaces, Hiroshima City University,3-4-1, Ozuka-higashi, Asaminami-ku, Hiroshima, 731-3194, JAPAN
国际会议
8th International Conference on Neural Information Processing(ICONIP 2001)(第八届国际神经信息处理大会)
上海
英文
629-634
2001-11-14（万方平台首次上网日期，不代表论文的发表时间）