With the vigorous development of intelligent city, it has been an important task that we need to consistently revitalize owned information, make full use of the incremental data, to provide unified data service for intelligent application in various fields. This paper proposes an improved power customer segmentation model, which uses an optimized algorithm that is based on the basic data of the intelligent city and the internal data of the electric power enterprise. This model attach importance to the customer's change tendency on the basis of maintaining the original clustering distance, making the result of clustered classification more detailed and more advanced. At the same time, the quality of clustering center is improved and the convergence rate of clustering is improved by using the pre-clustering and average clustering center linear prediction. The simulation results show that this model provide a better classification way for power customers, which has higher classification quality and faster speed.
[1]
Yi-Hui Liang,et al.
Integration of data mining technologies to analyze customer value for the automotive maintenance industry
,
2010,
Expert Syst. Appl..
[2]
Jie Lu,et al.
A New Index and Classification Approach for Load Pattern Analysis of Large Electricity Customers
,
2012,
IEEE Transactions on Power Systems.
[3]
Paul D. McNicholas,et al.
Model-based clustering, classification, and discriminant analysis via mixtures of multivariate t-distributions
,
2011,
Statistics and Computing.
[4]
Jianwei Guo,et al.
Research on Distributed Data Mining System Based on Hadoop Platform
,
2014
.