论文部分内容阅读
Freeman在大量神经生理实验的基础上建立了生物嗅觉系统的非线性神经网络模型,即K系列模型.其中,KIII模型是一种混沌神经网络,它的模式识别机制与以往的人工神经网络完全不同,更接近实际生物神经系统的工作模式.研究通过对26个英文字符的学习、识别研究得出,系统相对于传统的神经网络有着很强的学习能力,学习5～6次就能有很好的识别能力,在10次达到最优学习效果,并与真实神经系统学习过程中的倒“U”曲线相对应.
Freeman has established a nonlinear neural network model of biological olfactory system based on a large number of neurophysiological experiments, namely the K series model, in which the KIII model is a chaotic neural network whose pattern recognition mechanism is completely different from the previous artificial neural network , Closer to the working model of the actual biological nervous system.Study on learning and recognition of 26 English characters has shown that the system has a strong learning ability compared with the traditional neural network and can be well learned 5 to 6 times Of the recognition ability, in the 10 to achieve the best learning effect, and with the real nervous system learning process inverted “U ” curve corresponding.