Customer Churn Prediction
Problem: Train a machine learning model to predict customer churn for a telecommunications company.
Context: Customer churn is a big problem for organizations in every industry. With data analytics and machine learning, we can identify factors that lead to customer turnover, create customer retention plans, and predict which customers are likely to churn.
Model type: Logistic regression
What we did: We trained a logistic regression model on a dataset from a telecommunications company (dataset can be found here). We trained the model using SnapLogic Data Science, a self-service solution for the entire machine learning lifecycle. Our logistic regression model predicts the probability of customer churn with an accuracy of 80.58 percent on average. (More on how we built this demo.)
In the bar chart to the right, you can see that most of the churning customers subscribed to the fiber optic internet service. If you select "Contract" from the drop-down filter, you can see that most of the churning customers were on month-to-month contracts as opposed to one-year or two-year plans. Explore the drop-down filter to see which factors contributed to customer churn.
The table below contains information on 10 customers from the dataset. The predictions are in the "Churn" column (scroll to the right). Try changing the data to get new predictions in real-time.