Tehran House Price Forecasting Project

Project Structure

• data/input/tehranhouses.csv: Main dataset of Tehran houses
• data/output/best_stacking_model.pkl: Trained and saved model
• notebooks/tehran.ipynb: Main notebook containing steps for analysis, data cleaning, feature engineering, modeling, and evaluation
• images/: Output images and visualizations

Project Workflow

1. Business Understanding and Objective Definition
2. Data Collection and Exploration
3. Data Cleaning and Preprocessing
4. Exploratory Data Analysis (EDA) with Plotly
5. Feature Engineering and Dimensionality Reduction
6. Modeling with Various Regression Algorithms
7. Model Evaluation and Selection of the Best Model
8. Saving the Final Model

How to Run

1. Install the required packages:

   pip install -r requirements.txt

2. Run the notebook notebooks/tehran.ipynb and follow the steps.
3. The final model is saved at data/output/best_stacking_model.pkl.

Best Model

The final model is a Stacking Ensemble Regressor, combining XGBoost, Gradient Boosting, and Random Forest with a Ridge meta-learner, achieving high predictive accuracy.

License

This project is licensed under the LICENSE.