| Abstrakt: |
With their digital marketing efforts, banks are now attempting to address the needs of their existing clients. By storing large-scale data gathered from marketing studies, it is known to provide statistical results in order to forecast client behaviour in artificial intelligence applications. Using real banking marketing data, including customer profiles, this study compared the performance of random forest and logistic regression methods. These algorithms were also tested on the WEKA, Google Colab, and MATLAB platforms to compare performance. At the conclusion of the trial, the random forest algorithm on the WEKA platform produced the best results, with 94.8 percent accuracy, 93.9 percent sensitivity, 94.8 percent recall, 94.4 percent f1-score, and 98.7 percent AUC value. Furthermore, when compared to previous investigations, the obtained performance values generate superior outcomes. [ABSTRACT FROM AUTHOR] |
