感知准则函数是五十年代由Rosenblatt提出的一种自学习判别函数生成方法，Rosenblatt企图将其用于脑模型感知器，因此被称为感知准则函数。其特点是随意确定的判别函数初始值，在对样本分类训练过程中逐步修正直至最终确定。

-Perception criterion function is the 1950s proposed by Rosenblatt, a self-learning discriminant function generation method, Rosenblatt attempt for a brain model of perception, is known as the perception of the criterion function. Its characteristics are free to determine the discriminant function of the initial value, the gradual modification in the sample classification training process until finalized.