Chaos and neural networks
Chaos and neural networks
复制标题
混沌和神经网络
DOI：
10.1007/978-1-4757-9631-5_23
复制
发表时间：
1987
期刊：
影响因子：
--
通讯作者：
E. Lábos
中科院分区：
文献类型：
--
作者：
E. Lábos
神经系统是由神经元模块组成的紧密连接的网络。不规则或周期性的神经自活动可能来源于单元或网络。这里提出的网络的正式概念是由有限数量的单元或组成变量（细胞）组成的可分离系统。这些“状态变量”相互作用，类似于神经细胞。这意味着它们的未来是由一系列其他变量的过去历史决定的。在模型中，不同的形式变量集可能对应于真实的单元（细胞）或真实的神经元网络。在特定情况下，规则（稳定）单元在连接到网中时可能变得不规则或不稳定。在其他例子中，不稳定或不规则的单元活动可能转变为稳定或周期性的功能。在某些情况下，预测网络中互联单元的命运是可能的。然而，目前还没有关于相互联系可能产生的后果的一般理论。
Nervous systems are strongly connected networks of building modules the neurons. The irregular or periodic neural autoactivity might take its origin either from units or networks.The formal concept of network presented here is a separable system of a finite number of units or component variables (cells). These ‘state-variables’ interact, similarly to the nerve cells. This means that their future is determined by the past history of a set of other variables. In models the different sets of formal variables may correspond either to real units (cells) or real networks of neurons.In specific cases regular (stable) units may become irregular or unstable when coupled into nets. In other examples unstable or irregular unit activities may turn into stable or periodic functioning. The prediction of the fate of the interconnected units in a network in some cases is possible. Nevertheless, no general theory of the possible consequences of interconnections is now available.