House Price Prediction (Regression) 🏠

📌 Project Objective

The goal is to build a robust regression model capable of predicting the continuous variable SalePrice based on 79 explanatory variables describing almost every aspect of residential homes.

🛠️ Key Workflow & Concepts

• Target Variable Analysis: Identified positive skewness in SalePrice and applied Log Transformation (np.log1p) to normalize the distribution for better model performance.
• Advanced Data Preprocessing:
  • Handled missing values using context-aware strategies (e.g., filling LotFrontage with the median of the neighborhood).
  • Conducted feature encoding using Label Encoding for ordinal data and One-Hot Encoding for nominal data.
• Feature Engineering: Scaled numerical features using StandardScaler to ensure uniform influence on the model.
• Model Comparison: * Built a baseline Linear Regression model.
  • Implemented an advanced XGBoost Regressor to handle non-linear relationships and improve accuracy.
• Evaluation Metrics: Evaluated models using RMSE (Root Mean Squared Error), MAE (Mean Absolute Error), and R-squared.

📊 Results Summary

• Skewness Correction: Reduced SalePrice skewness from 1.88 to a more normal distribution.
• Model Performance: The XGBoost model significantly outperformed the baseline, capturing complex feature interactions that linear models missed.

🚀 Repository Structure

Project-House_Price_Prediction/
├── Project_House_Price_Prediction.ipynb   # Full ML Pipeline
├── train.csv                              # Training Data
├── test.csv                               # Testing Data
└── README.md                              # Project documentation