This project leverages ensemble learning techniques to predict house prices based on the Kaggle House Prices - Advanced Regression Techniques dataset.
-
Data Preprocessing
- Duplicate Handling
- Missing Value Imputation
- Encoding Categorical Features
- Feature Scaling
- Log Transformation
- Feature Selection via Correlation Analysis
-
Ensemble Modeling
- Bagging Models:
- Random Forest Regressor
- Extra Trees Regressor
- Boosting Models:
- Gradient Boosting Regressor
- XGBoost
- Bagging Models:
-
Hyperparameter Tuning
- Conducted using Optuna, a powerful framework for automated hyperparameter optimization.
- Python
- Libraries:
pandas,numpy,scikit-learn,matplotlib,optuna,xgboost - Streamlit: For building an interactive app
Experience the model in action via a Streamlit-powered interactive demo:
The app presents a randomly generated house profile with its features. You’ll see the model’s predicted price and be asked:
“Is the real price higher or lower?”
Can you beat the model?
- Kaggle House Prices Dataset:
https://www.kaggle.com/c/house-prices-advanced-regression-techniques