Predicting the accuracy for telemarketing process in banks using data mining

Abstract EN

Managing and analyzing the data that generates from routine business operations is one of the biggest challenges for the banking industry.

Banks play a significant role in the growth of the economy as well as provide numerous financial services to the customers.

In those services, telemarketing is a common business strategy applies in the bank to offer and promote new products and services to its customers.

This type of campaign produces very large dataset, proper analysis of those data can support the bank in planning future strategies.

Therefore, this study proposed a data mining approach, to analyze and predict using the telemarketing campaign dataset.

The dataset was prepared based on the pieces of evidence collected from the customers, during the live call session organized by the bank.

To implement the proposed model, we selected the real dataset generated during the telemarketing process.

Logistic Regression, Decision Tree, and Multilayer Perceptron were part of conducted experiments.

The cross-validation strategy applied for measuring and comparing the performances of each algorithm.

The result suggested that logistic regression provides the best accuracy among the three models, recorded as 91.48%.

The research is helpful for the banking industry, where the model can be used to predict the success of telemarketing calls and understand the patterns within the dataset.

Moreover, the decision-makers can use the model for defining their future strategies to run the telemarketing process efficiently.