To Predict Customer Churn By Using Different Algorithms

Customer churn prediction is the supreme thing for every company. Every company is facing that issue. A company can easily predict how many customers will stay and take their service and how many customers will omit their service. This paper predicts if their customers will take their service in the future or not. In our paper, we have used the " Impact Learning" algorithm to predict customer churn. The data is trained by the impacts of features from the intrinsic rate of natural increase in the impact learning algorithm. Here, the training set of data is numerical data. Our proposed model is implemented by using three stages namely data collection, identifying null value, and data preprocessing. This paper has also shown the performance comparison between Logistic regression, Impact learning, and Artificial Neural Network algorithm. From the comparison, we can see that the Impact learning algorithm gives the best and higher accuracy than the other two algorithms.