The bifurcating neuron network 3
The bifurcating neuron network 3
复制标题
分叉神经元网络 3
DOI：
10.1016/s0893-6080(00)00083-6
复制
发表时间：
2005
期刊：
影响因子：
--
通讯作者：
N. Farhat
中科院分区：
文献类型：
--
作者：
Geehyuk Lee;N. Farhat
分叉神经元（BN）是一种混沌积分激发神经元，是一种通过来自环境的相干调制增强的神经元模型。BN在数学上等价于正弦圆映射，这种等价关系使我们能够将一维映射的数学应用于BN网络的设计。BN的对称性研究表明，BN可以被配置为表现出由吸引子合并危机控制的双稳性。此外，双稳态的对称性，可以通过引入一个正弦波动的阈值电平的BN控制。这两个观察结果使我们设计了BN网络1（BNN-1），这是一种表现出联想记忆的混沌脉冲耦合神经网络。在数值模拟中，BNN-1表现出更好的性能比连续时间Hopfield网络，至于spurious极小值问题的关注，并表现出许多生物合理的特性。
The Bifurcating Neuron (BN), a chaotic integrate-and-fire neuron, is a model of a neuron augmented by coherent modulation from its environment. The BN is mathematically equivalent to the sine-circle map, and this equivalence relationship allowed us to apply the mathematics of one-dimensional maps to the design of BN networks. The study of symmetry in the BN revealed that the BN can be configured to exhibit bistability that is controlled by attractor-merging crisis. Also, the symmetry of the bistability can be controlled by the introduction of a sinusoidal fluctuation to the threshold level of the BN. These two observations led us to the design of the BN Network 1 (BNN-1), a chaotic pulse-coupled neural network exhibiting associative memory. In numerical simulations, the BNN-1 showed a better performance than the continuous-time Hopfield network, as far as the spurious-minima problem is concerned and exhibited many biologically plausible characteristics.
DOI：
10.1016/0013-4694(80)90319-3
发表时间：
1980-01-01
期刊：
ELECTROENCEPHALOGRAPHY AND CLINICAL NEUROPHYSIOLOGY
影响因子：
--
作者：
BRESSLER, SL;FREEMAN, WJ
通讯作者：
FREEMAN, WJ