Application Of Stacking Integrated Model In Bank Telemarketing

With the deepening of social informatization,the financial field,where new technology has always had grand prospects,is constantly updating its operation and management modes.Banking industry,relatively more mature than other sectors in financial field,taking advantage of its convenience of obtaining a large number of customer data,can analyze unique intrinsic value from customer information.Faced with massive data sets,any bank that wants to improve its core competitiveness cannot avoid this issue,that is,how to use data mining technology to achieve long-term business breakthroughs in existing situations.This article is based on the data collected from a bank's telemarketing campaign to predict whether customers will buy the products of this bank.First of all,the background,significance,research status at home and abroad and research methods and ideas of this issue are introduced.Second,relevant theories of bank telemarketing under the background of big data technology and the data mining methods adopted in this paper are described.Third,the pre-processing of the empirical data is carried out,andthree traditional learning classifiers,CART decision tree model,logistic regression model and support vector machine model,are established.The Stacking-Adaboost integrated classifier is designed by combining Stacking integration and Adaboost algorithm and the generalization ability of each classifier is analyzed and evaluated by several indicators.Finally,based on the predicted results of the model,marketing suggestions for bank telemarketing campaign are provided to maximize the use of existing resources and improve the effective value of marketing activities.