Association rules algorithm in bank risk assessment

Abstract

Domestic financial enterprise data management applications generally brings together the vast amounts of data, but can not find the relationship and business rules exists in the data to do risk prediction assessment. Therefore, the domestic financial companies need to accelerate the pace of information technology in regions of integration of customer resources, business analysis and investment decisions. This paper analyzes the risk assessment approach of banks, mainly focuses on the analysis of association rules data mining in bank risk assessment, and discusses the working principle of improved association rules algorithm genetic algorithm in commercial bank risk assessment. We described the methods and processes of system application. We select the matrix form, only scan the database once, and use the method of selecting assumption frequent items and numbers, find the frequent item sets through high end item sets, minimize the number of candidate data sets, greatly improve the efficiency of the algorithm. © 2011 Springer-Verlag Berlin Heidelberg.