一种优化 BP 神经网络训练样本的方法

BP神经网络训练样本的选取对网络的泛化能力有较大的影响,特别,怎样从高维大样本数据中选取合适训练样本是一个难点。本文运用因子分析法对大样本数据进行预处理,再利用分析所得的公因子进行聚类分析,这样既可以降低指标的维数,也可以减少样本的数量。实验证明,该方法简化了网络结构、加快了网络的收敛速度,对提高网络的泛化能力有一定的帮助。

An Approach to Optimizes the Training Samples of BP Neural Network

The selection of BP neural network training samples has a strong impact on the generalization ability of the network,and the selection of training samples from the high-dimensional data is especially difficult.The paper uses the method of factor analysis to pretreat the data from a large samples,and then cluster the result.In this way,both of the dimensions of variables and the number of samples can be reduced.The result of the experiment indicates that the method can improve the convergence speed and simplify the network structure as well as improve the network generalization.