摘要
In this paper a new associative memory neural network is proposed. Neuron state is a vector. A pattern stored in the network with N neurons consists of N second level patterns with M components. Each pattern is stored in a "pattern loop" which is composed of N links. A link consists of "link state" and "inhibited path", the former is used to store second level pattern, and the later to eliminate spurious pattern. Its storage capacity can be up to (N - 1)!; it can completely suppress spurious patterns and permit to input incomplete patterns for recalling; and can recall all or number of the stored patterns which have the minimum Hamming distance with input pattern.
- 具有二级模式的多状态神经元联想记忆网络王剑毛宗源计算机工程, 2004, (04): 16.
- Analysis and Design of Multivalued High-Capacity Associative Memories Based on Delayed Recurrent Neural NetworksZhang, Jiahui; Zhu, Song*; Bao, Gang; Liu, Xiaoyang; Wen, ShipingIEEE Transactions on Cybernetics, 2022, 52(12): 12989-13000.被引 9
- 一种以环路作为存储单元的联想记忆神经网络王剑毛宗源计算机工程与应用, 2003, (04): 26.
