BCG Data Science Project: Customer Churn Prediction Overview This project is a part of a free program on Forage, designed to be completed in approximately 1-2 days. It focuses on predicting customer churn for a gas and electricity utility company using a variety of data science techniques.
Objectives Exploratory Data Analysis (EDA): Conduct initial investigations on the data to discover patterns, spot anomalies, frame hypotheses, and check assumptions. Data Cleaning: Prepare the raw data by cleaning and structuring it for the analysis. Feature Engineering: Develop features from the data that are suitable for predictive modeling. Predictive Modeling: Use a Random Forest classifier to predict customer churn. Tools and Technologies Python: Primary programming language used for data analysis and modeling. Pandas & NumPy: For data manipulation and numerical computation. Scikit-learn: For implementing machine learning algorithms. Matplotlib & Seaborn: For data visualization. Dataset The dataset includes historical customer data, pricing information, and churn indicators. Specific features include usage patterns, account details, and billing information.