Used_Car_Price_Prediction_ML_CaseStudy

Used Car Price Prediction – ML Data Preparation

Cleaned and engineered a 500-record used car CSV dataset for a machine learning price prediction model. Key work: standardized mixed Mileage formats ('224k' → 224000), removed 18 outlier/corrupt price rows, imputed 44 missing values using Median, applied Label & One-Hot Encoding, and derived a Car_Age feature (2024 – Year) as a stronger predictor. Final output: 482 rows × 18 model-ready features. Recommended model: XGBoost/Random Forest with Price as target.