该算法以核函数代替原始数据在特徵空间中映射值的内积，并且神经元权值向量的初始化和更新都可由其组合系数向量表示，从而获得了直观而简单的迭代公式。

该算法以核函数代替原始数据在特徵空间中映射值的内积，并且神经元权值向量的初始化和更新都可由其组合系数向量表示，从而获得了直观而简单的迭代公式。的英文翻译

基本释义

• The inner product of the mapping value of the original data in feature space is replaced by a kernel function, and the weights of each neuron can be initialized and updated by initializing and updating the combinatorial coefficient vector of each weight in the algorithm of KSOM, so some intuitive and simple iteration formulas are obtained.