SOM神经网络，Self-Organizing Map 的缩写，即自组织映射。

1981年芬兰Helsink大学的T.Kohonen教授提出一种自组织特征映射网，简称SOM网，又称Kohonen网。 Kohonen认为：一个神经网络接受外界输入模式时，将会分为不同的对应区域，各区域对输入模式具有不同的响应特征，而且这个过程是自动完成的。自组织特征映射正是根据这一看法提出来的，其特点与人脑的自组织特性相类似。

由于它的强大功能，多年来，神经网络在数据分类、知识获取、过程监测、故障识别等领域中得到广泛运用。-SOM neural network, Self-Organizing Map acronym, that self-organizing map. 1981 University Professor T.Kohonen Finland Helsink propose a self-organizing feature map network, referred to as SOM network, also known as Kohonen network. Kohonen think: when a neural network to outside input mode, will be divided into different corresponding regions, the regional input to the model with different response characteristics, and this process is done automatically. It is similar to the self-organizing map according to this view put forward, which is characterized by self-organization and phase characteristics of the human brain. Because of its powerful over the years, neural networks in data classification, knowledge acquisition, process monitoring, fault identification and other areas to be widely used.