Input and Output

Understanding how data moves from input to output helps clarify how raw telecom data becomes a predictive churn solution.

Input

• A telecom customer churn dataset in CSV format.
• The dataset includes information such as customer demographics, tenure, services subscribed, contract type, payment method, monthly charges, total charges, and churn status.
• The file is uploaded into Google Colab and loaded into a Pandas DataFrame for processing.

Output

• A cleaned and preprocessed dataset with missing values handled and categorical variables encoded.
• Engineered features that improve predictive performance.
• A trained Logistic Regression model capable of predicting churn probability.
• Model evaluation metrics including accuracy, precision, recall, F1-score, and confusion matrix.
• An interactive Gradio application that allows users to input customer details and receive churn probability predictions.
• Clear business insights that help identify high-risk customers and support proactive retention strategies.