Machine learning models analysis on customer churn data, for insides (key factors) and prediction Nowadays, telecom industry faces fierce competition in satisfying its customers. The role of churn prediction is not only to accurately predict churners but also to interpret customer churn behavior. The objectives of this project are using machine learning to predict and to find the root causes of churn. The dataset used is from kaggle. (https://www.kaggle.com/jpacse/datasets-for-churn-telecom)
After data exploratory analysis, applied careful feature engineering Various machine learning models are used to study costumer churn In-depth study with Logistic regression: a) ROC and CAP curves; b) key factors for churn Discoveries and recommendations are summarized.