Artificial immune network with feature selection for bank term deposit recommendation

Artificial immune systems (AIS) have been widely utilized for pattern recognition and data analysis in various fields of science and technology, and artificial immune networks (AIN) are based on AIS. In this study, an artificial immune network is used for collaborative filtering as a classification model for bank term deposit recommendations, once feature selection has been applied to filter out key features for classification purposes. AIN is used to represent a network of customers with bank term deposits, and it can be adopted as a group decision-making model in predicting whether a new customer will have a term deposit or not. Formulae for calculating the affinity between an antigen and an antibody, and the affinity of an antigen to an immune network are also developed. A series of experiments are conducted, and the results are very encouraging. Despite the class imbalance problem in the test dataset, the proposed model outperformed other models, achieving the highest accuracy in testing.