An AI-powered Real Estate Analytics Dashboard that predicts residential property prices across major Indian cities using Machine Learning.
The application combines data preprocessing, feature engineering, machine learning, interactive visualizations, and location intelligence to provide users with accurate property price estimates and valuable market insights.
Predict property prices based on:
- Property Location
- Area Name
- BHK Configuration
- Bathroom Count
- Balcony Count
- Property Size (sqft)
- Furnishing Status
- Facing Direction
- Floor Information
- Property Status
- Transaction Type
Compare predicted property value against market averages.
Displays:
- Average Market Price
- Predicted Property Price
- Percentage Difference
Automatically categorizes properties as:
- Budget Property 💰
- Premium Property ⭐
- Luxury Property 👑
Visualizes the most influential features affecting price prediction using Random Forest Feature Importance.
Examples:
- Area Name
- Area Size
- BHK
- Bathrooms
- City
- Floor Information
Displays selected city location on an interactive map.
Supported Cities:
- Hyderabad
- Mumbai
- Delhi
- Chennai
- Pune
Provides a quick overview of:
- City
- Area
- BHK
- Bathrooms
- Balcony
- Area Size
Dataset-level insights:
- Average Property Price
- Median Property Price
- Maximum Property Price
- Missing value handling
- Data formatting
- Feature preparation
Applied to:
- Status
- Transaction
- Furnishing
- Facing
- City
Applied to:
- Area_Name
StandardScaler applied to:
- Bathroom
- Balcony
- BHK
- Area(sqft)
- Floor_No
- Total_Floors
Used for final property price prediction due to its ability to capture complex relationships and non-linear patterns in real estate data.
ML_House_Price_Prediction/
│
├── app.py
├── ML_cleaned_real_estate.csv
├── Real_Estate_Price_Predictor.pkl
├── requirements.txt
├── README.md
│
└── Screenshots/
├── Dashboard.png
├── prediction.png
└── demo.mp4
- Streamlit
- Pandas
- NumPy
- Scikit-Learn
- Category Encoders
- Joblib
- Streamlit Charts
- PyDeck Maps
| Metric | Value |
|---|---|
| Cities | 5 |
| Records | 2877 |
| Features | 12 |
| Model | Random Forest |
Supported Cities:
- Delhi
- Mumbai
- Chennai
- Hyderabad
- Pune
Four regression algorithms were evaluated to identify the most suitable model for property price prediction.
| Model | R² Score | Adjusted R² |
|---|---|---|
| Random Forest Regressor | 0.5745 | 0.5536 |
| Decision Tree Regressor | 0.5324 | 0.5094 |
| KNN Regressor | 0.4155 | 0.3867 |
| Linear Regression | -1.9551 | -2.1007 |
Random Forest Regressor achieved the highest R² and Adjusted R² scores among all evaluated models and was therefore selected as the final deployment model.
The model captures approximately 57% of the variance in property prices, outperforming Decision Tree, KNN, and Linear Regression models on the dataset.
Clone the repository:
git clone https://github.com/yourusername/ML_House_Price_Prediction.gitMove into project directory:
cd ML_House_Price_PredictionInstall dependencies:
pip install -r requirements.txtRun the application:
streamlit run app.py- Hyperparameter Optimization
- Property Recommendation System
- Price trend forecasting
- Cloud deployment with Streamlit Cloud or AWS
- Interactive City-Level Dashboards
- Model Performance Monitoring
- Advanced location intelligence using GIS
Aspiring Data Analyst & Machine Learning Engineer