Abstract:

Total of two data sets which is published from kaggle.com has been used in this study. The first data set belongs bank customers, and the second data set contains telephone operator customers. Customer loss analysis was conducted by examining various attributes with using the two data sets. In this study, Logistic Regression, Naive Bayes, Desicion Tree, K-NN, SVM and LDA classification modeling are used. The methods which are determined from 13.000 customer information are used in this research. Python Programming Language, which is easy to implement and has many literature studies, has been chosen. The classification methods used were applied for both data sets and their accuracy rates were compared. It has been observed that the most decisive result is the decision trees algorithm. More than 70% of all results give an accuracy rate, a successful study is revealed.

Keywords: