Financial Product Recommendation Research Based On Artificial Immune System Combines Collaborative Filtering For Bank Term Deposit

Transparency within the financial industries and the multiple choices of financial products is increasing competition in the financial market. The ability to predict the preference of customer is very important to every financial service company, and to developing a high-efficient classification model can make a company increasing profit and reducing the cost. So, the main idea of this research presents an artificial immune classification combines collaborative filtering approach to predict the customer will subscribe a term deposit or not. Artificial immune systems(AIS) are intelligent systems inspired by the processes of the vertebrate immune system. AIS are widely used in biologically-inspired computing, natural computation, and machine learning. Collaborative filtering(CF) has become a very popular technology in information filtering and information system in recent years. And now widely used in various business activities, such as Amazon, MovieFinder, Taobao, and so on, and it also has a remarkable effect. But there are still some problems in CF, this research combines AIS to solve the cold start problem. To verify the proposed model is effective, this study considers real data collected from a Portuguese retail bank to test and compare with decision trees(DT), support vector machine(SVM), logistic, na?ve Bayes(NB), artificial immune recognition system(AIRS1) and immunos99. The experimental results show that AIS performs much better than DT and SVM.