Research And Application Of Data Mining Technology In Bank Precise Marketing

With the rapid development of computer technology,banks have also made rapid progress in the level of informationization.With continuous improvement of their businesses and continuous improvement of intelligence,banks have produced a large amount of business data such as customer attribute data,Intermediate data and unstructured data and other information.Due to the current severe economic situation,the intense competition within the industry and the impact of Internet finance on banks,commercial banks need to find a more accurate marketing solution so as to provide banks with accurate decisions.Data mining technology is to extract valuable information,which can not be captured by human beings manually,from the massive data,so as to provide a reference for the bank's decision-making.Data mining is only based on the data itself,data cleaning,change,and data mining model set up to dig out in a large number of potential data in the decision-making.This article addresses the data obtained from direct marketing activities of banking institutions.The main work done is:(1)Describe the current development of commercial banks and the development of customer relationship management mode.At the same time,combine the bank's informationization with data mining technology,explain the breakthrough point,how to combine the data of bank system and data mining technology to get the data In the hidden value of the information.(2)In this paper,we get the data set about the successful purchase of wealth management products by the users through the telemarketing scheme and build the model based on this data set.(3)Data preprocessing,such as data discretization and standardization,is performed on the acquired data.This process,as a pretreatment process of data mining,plays a key role in model building(4)The SMOTE algorithm is used to eliminate the imbalance between the two types ofsamples.At the same time,genetic algorithm is used to filter out the redundant features.Finally,a decision tree classification model is adopted,and the correct result is identified on the test set.