Marziyeh Ajorloo - MSc. Student, Dept. of Computer Engineering Damavand Branch, Islamic Azad University Tehran, Iran
Fatemeh Safara - Faculty Member: Dept. of Faculty Member Islamshahr Branch, Islamic Azad University Tehran, Iran

Customer Relationship Management (CRM) is a term for the strategies, methodologies and tools that help businesses in managing customer relationships in an effective way. Data Mining has an important role in extracting the detail information from a large number of customer databases. By using MLP (Multi-Layer Perceptron) Neural Networks for CRM, companies and organizations can achieved advantages in business. This paper is to study the effectiveness of MLP on the bank direct marketing and two classification models, Logistic Regression (LR) and MLP are found effective. The experimental results show that MLP and LR achieved an accuracy of 95.70% and 91.06% respectively, where MLP was more accurate than LR in predicting customer behaviors.

Customer Relationship Management; Bank Direct Marketing; Data Mining; Multi-Layer Perceptron Neural Network; Logistic Regression