1
Exploratory Data Analysis
Master pandas and seaborn to uncover trends, seasonal patterns, and feature correlations in real automotive data.
Build ML models to predict resale values for VW Golf and Audi A4 using real UK market data. Compare Linear Regression with Random Forest and deliver business recommendations in this simulated case study.
Just completed
Engineered an ML pricing engine in Python comparing Linear Regression and Random Forest models on UK used-car listings. Delivered risk-assessment recommendations to optimize CPO vehicle pricing. Graded and verified by ProoV.
Get introduced to the ProoV platform, the Volkswagen Group case study, and the core problem: predicting used car resale values using real UK market data.
Load VW and Audi datasets. Merge into a single DataFrame. Perform Exploratory Data Analysis (EDA) including correlation heatmaps, boxplots by brand, and scatter plots comparing price to mileage and year.
Engineer derived features like car age and mileage per year to capture usage intensity. Handle outliers and one-hot encode categorical variables like transmission and fuel type.
Establish a Linear Regression baseline, then train a RandomForestRegressor. Compare R², MAE, and RMSE. Plot feature importances to discover the strongest price predictors.
Interact with the in-app Price Predictor to validate model intuitions. Write an executive summary for the CPO team outlining the winning model, top features, and specific price insights.
Master pandas and seaborn to uncover trends, seasonal patterns, and feature correlations in real automotive data.
Transform raw data into powerful predictive signals. Handle missing values, outliers, and categorical encoding.
Train and evaluate Scikit-Learn models like Random Forests to predict precise vehicle resale values.