Clustering Algorithm Based on User Power Load

Abstract:
Cluster analysis is a very important technology and method in data mining, and it is one of the main tasks of data mining. Cluster analysis is to use clustering algorithm to find meaningful classes, mainly based on the similar samples are divided into a class, and the samples with large differences are distinguished, so that the generated cluster is a set of data objects, these objects are similar to the objects in the same cluster, but different from the objects in other clusters. In applications, data objects in a cluster are often treated as a whole. Under the background of China's power grid companies to strengthen the reform of demand side and more effectively carry out economic dispatching of power system osperation, this paper focuses on the study of several classic clustering algorithms at present, analyzes the principle and compares the performance, and expects to choose the most practical method when there is a specific demand.