Data Dimensionality Reduction Method for Smart Meter Big Data Analysis
Aiming at the problem of excessive computational complexity in the process of smart meter big data cluster analysis,a data dimensionality reduction method based on polar coordinate projection is proposed.The polar coordinate projection meth-od is used to map the data set composed of total power consumption and peak power consumption into a projection matrix com-posed of polar angle and polar diameter,and a data set composed of total power consumption and peak power consumption is designed.The improved clustering distance metric is used to determine the optimal number of clusters using the sum of squared error method.This method is introduced into the k-means algorithm for example analysis.The results show that the dimen-sionality reduction method can effectively improve the performance of the clustering algorithm and reduce the computational complexity of the clustering algorithm.
万方数据
维普
