Comparative study of the classification models for prediction of bank telemarketing

Elzhan Zeinulla, Karina Bekbayeva, Adnan Yazıcı

2018

Closed access Metadata (CSL-JSON)
DOI 10.1109/icaict.2018.8747086
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

This research paper has evaluated various classification models for prediction of bank telemarketing campaign results regarding the probability of the subscription of the customer to the deposit. The effectiveness of these algorithms has been evaluated by Receiving Operator Characteristic (ROC) and Cumulative Accuracy Profile (CAP) curve analysis, the accuracy of the algorithm and variance of the predictions. According to the results of the research, the best model for bank telemarketing effectiveness prediction are Random Forest and Deep Artificial Neural Network. The Logistic Regression and Naive Bayes are not as suitable as the other classification methods for this kind of problems due to the poor accuracy and overfitting issues.