Prediction of Term Deposit in Bank: Using Logistic Model

Telemarketing remaining to be one of the most popular direct marketing techniques, which is urged make accurate predictions. It critical for banks have a mature machine learning approach forecasting problems. Through analysis time deposit telemarketing data Portuguese commercial on Kaggle.com, bank can more accurately locate target customers, so as improve efficiency increasing amount business. Firstly, are preprocessed, coded and shuffled divide test set training set. Secondly, logistic regression carried out eliminate irrelevant independent variables through correlation analysis, makes prediction accurate. The results were compared with decision tree retrograde comparison. Accuracy Area Under Curve (AUC) used evaluation indexes. demonstrate effectiveness model elimination variables. This method applied actual problems find customers accuracy marketing.