Research and application of subtilized customer clustering algorithm in power marketing

Customer oriented diversification service lacks effective and accurate grouping under the background of the new electricity reform. In this paper, a logic regression model based on probability clustering is proposed to realize the refinement of the grouping of arrears risk in the electricity risk prevention and control business. Through the analysis and positioning of customer groups, refining, induction, and segment customer characteristics, it can achieve the purpose of customer refinement. And then develop a precise differentiation service strategies and effect evaluation method. Example analysis shows that the practical conditions of the model are already available, and the model can be used as an important support for high quality service and lean management of power grid.