Wieding J U, Schönle P W
Abteilung Klinische Neurophysiologie, Universität Göttingen.
Nervenarzt. 1991 Jul;62(7):415-22.
Neural networks are used as models of cognitive systems. On the basis of results from brain research, the processing of knowledge and information can be conceived as occurring through parallel interaction of multiple but simple and uniform processor elements in a network structure. In these 'neural networks' knowledge is stored in a distributed way throughout the network and subjected to parallel processing. According to Hebb's concept of synaptic plasticity, the process of learning is based on the strengthening of the links between the processing elements. The implementation of network models in electronic data processing systems allows for simulation of cognitive phenomena such as learning and forgetting or abilities such as recognition of acoustical or optical patterns, in a more efficient way than conventional computer programming. Neural network modelling can help to understand the abilities and disorders of the brain.
神经网络被用作认知系统的模型。基于大脑研究的结果，知识和信息的处理可以被设想为通过网络结构中多个但简单且统一的处理器元件的并行交互来进行。在这些“神经网络”中，知识以分布式方式存储在整个网络中，并进行并行处理。根据赫布的突触可塑性概念，学习过程基于处理元件之间连接的强化。在电子数据处理系统中实现网络模型，能够比传统计算机编程更有效地模拟诸如学习和遗忘等认知现象，或诸如识别声学或光学模式等能力。神经网络建模有助于理解大脑的能力和障碍。
