This project predicts the Fire Weather Index (FWI) using meteorological and spatial features from the Algerian Forest Fire Dataset. The model is trained using Ridge Regression, a linear regression technique with L2 regularization, to predict fire risk levels.
- Uses Algerian Forest Fire Dataset (UCI Repository)
- Takes user inputs such as:
- Temperature (°C)
- RH (Relative Humidity)
- WS (Wind Speed)
- Rainfall (mm)
- Region (Bejaia or Sidi-Bel Abbes)
- Fine Fuel Moisture Code
- Duff Moisture Code
- Initial Spread index
- Ridge Regression model for prediction of Fire Weather Index (FWI)
- Preprocessing includes feature engineering & one-hot encoding
- Python 3.x
- Pandas, NumPy, Scikit-learn, Matplotlib, Seaborn
- Ridge Regression (from
sklearn.linear_model)
Algerian Forest Fire Dataset: UCI Machine Learning Repository
- Clone this repository:
git clone https://github.com/your-username/your-repo-name.git cd your-repo-name
2.(Optional)Create a Virtual Environment:-
python -m venv venvActivate it
- Install dependencies:
pip install -r requirements.txt
- Run the Flask App:
🛜BY default, the app runs on http://localhost:5000
python application.py
- Open in browser Visit http://127.0.0.1:5000/predictdata
Input:
- Temperature: 29°C
- RH: 57%
- WS: 18 km/h
- Rain: 0 mm
- Region: Sidi-Bel Abbes
- FFMC:-65.67
- DMC-3.4
- ISI-1.3
- Classes:- Fire
Output:
- Predicted FWI: 0.7754016798136201 (moderate fire danger)
This project was created as part of an educational tutorial led by Krish Naik. It closely follows the model structure and methodology taught in the original video.
This project is created purely for educational purposes as part of a tutorial by Krish Naik.
The codebase and dataset usage closely follow the original video content.
All credits for the model structure and preprocessing steps go to the original author.
No commercial use intended.