Informationization construction in electric power has been developed in recent years. Meanwhile, data of power plants cannot be well utilized to support decision. Clustering analysis, as an important embranchment of data mining, should be applied to electric power industry, which will enhance market competitiveness of power plants. After comparing and analyzing existing clustering algorithms, this paper expatiates the algorithm ideas, implement and improvement of DBSCAN, and improved algorithm is applied to electric power marketing analysis. Experiments show that the dense and sparse areas will be identified; the global distributing pattern and the attributes'correlativity will be found. Decision-makers will dig out different client groups from basic database and start analyzing. Then, the aim of decision support should be achieved.
|