<ul><li>Customer churn represents more than just a numerical metric — it’s a critical indicator of business health in the telecom industry.</li><li>The primary task was analyzing customer data to identify patterns that lead to churn and building predictive models to mitigate this issue.</li><li>Data manipulation tasks involved extracting demographics, internet service types, senior female customers, and focusing on short tenure or low total charges.</li><li>Predictive models were built using Keras, including a basic binary classification model, a model with dropout layers to combat overfitting, and a multi-feature classification model.</li></ul>

Customer Churn Prediction Using Keras

Discover more