基于低方差滤波算法的改进降维算法

Improved Dimensionality Reduction Algorithm Based on Low Variance Filtering Algorithm

数据预处理在深度学习中越来越重要,而对于处理高维数据,降维成了数据预处理过程中必不可少的一部分。低方差滤波算法是一种常用的降维算法,但降维前的归一化操作处理时间消耗一定的时间和计算资源。所以,本文提出了一种基于低方差滤波算法的改进降维算法,通过对方差取算数平方根再除以均值的方法,省去了在降维前的归一化操作,从而减少了数据预处理的计算时间,提高了算法的效率,节省了计算资源。

Data preprocessing is becoming more and more important in deep learning,and for processing high-dimensional data,dimensionality reduction has become an essential part of the data preprocessing process.The low-variance filtering algorithm is a commonly used dimensionality reduction algorithm,but the processing time of the normalization operation before dimensionality reduction consumes a certain amount of time and computing resources.Therefore,this paper proposes an improved dimensionality reduction algorithm based on a lowvariance filtering algorithm.By taking the square root of the variance and dividing by the mean,the normalization operation before dimensionality reduction is omitted.This reduces the calculation time for data preprocessing,improves the efficiency of the algorithm,and saves computing resources.