基于轴突信号理论的神经网络聚类算法

通过借鉴Raju Metherate提出的只有部分脑细胞发出的信号到达了大脑皮层的理论和Stephen R Williams提出的突触信号强度随着离神经细胞主体的距离的加大而减弱的理论，提出了基于轴突信号理论的神经网络聚类算法。此算法在较高维空间中具备和传统竞争神经网络相当甚至更高的聚类准确率；通过对神经网络训练结果的进一步分析可以作为主因素分析和空间降维处理的依据；通过对竞争层神经元之间权重的修正得到类别的自组织关系。最后通过实验证明算法的有效性。

Axon signal theory-based clustering algorithm of neural network

With reference to the theory that only a part of signal from brain cells can reach pallium put forward by Raju Metherate,and the theory that axon signal strength is reduced with distance increment from main body of neural cells raised by Stephen R Williams,axon signal theory-based clustering algorithm of neural network is presented in this paper.This algorithm possesses equivalent to and even higher clustering accuracy than traditional competitive neural network in space with higher dimension.The further analysis of training result of neural network can be seen as a basis of space dimension reduction and primary component analysis,and self-organization relationships of categories can thus be yielded by weights of neural neurons in competitive layers.Finally the effectiveness of this algorithm is proved in experiments.