User power consumption behavior analysis method based on feature optimization and auxiliary clustering

The invention provides a user power consumption behavior analysis method based on feature optimization and auxiliary clustering, and the method comprises the following steps: carrying out the conversion from high-dimensional data to low-dimensional data based on UMAP data dimension reduction, and generating a data set containing dimension reduction features; inputting user power consumption data as a feature vector, constructing a feature index, endowing a weight by using a CRITIC weight method, and combining with the data set to obtain an optimized feature set; and based on the optimized feature set, user behavior analysis is carried out through a k-means clustering method optimized by a sparrow algorithm. According to the method, k-means clustering is improved based on the sparrow search algorithm, the clustering center is optimized, the clustering performance is improved, the reasonability of user classification is further improved, the user classification efficiency is improved, and different types of pow