sangeeth-pogula/Algerian-fire-prediction

# Algeria Fire prediction In this project, I experimented with the step-by-step implementation of an ML project lifecycle. It involves data cleaning, exploratory data analysis (EDA), feature engineering, model training, and hyperparameter tuning. After completing the machine learning part, I developed a user interface using Flask and hosted the project on GitHub. The machine learning models I worked with include Linear Regression, Ridge, Lasso, and ElasticNet. I fine-tuned these models using RidgeCV, LassoCV, and ElasticNetCV, and found that LassoCV gave the best performance. ## Table of Contents - Installation - Usage - File Structure - API Endpoints - Logging - Error Handling ## Installation ### Prerequisites Ensure you have the following installed: - Python 3.7+ - Flask - Scikit-learn - Pandas - Numpy ### Setup 1. Install dependencies: ```bash pip install -r requirements.txt ``` 2. Ensure the necessary model and preprocessor files are available in the correct directories: - `Preprocessors/KNNImputer.pkl` - `Preprocessors/outlier_bounds.json` - `Preprocessors/RobustScalar.pkl` - `Model/LassoCV.pkl` 3. Run the Flask application: ```bash python application.py ``` ## Usage ### API Endpoints #### 1. `GET /` Returns the HTML form where users can manually input meteorological data to get predictions. #### 2. `POST /predict` Accepts JSON data in the following format to return a prediction: ```json { "region": "0", "Temperature": 29, "RH": 57, "Ws": 18, "Rain": 0.0, "FFMC": 65.7, "DMC": 3.4, "ISI": 1.3, "Classes": 0 } ``` **Response**: ```json { "prediction": 0.69 } ``` ### Example of a POST request using `curl`: ```bash curl -X POST 127.0.0.1 \\ -H "Content-Type: application/json" \\ -d '{ "region": "0", "Temperature": 39, "RH": 57, "Ws": 18, "Rain": 0.0, "FFMC": 65.7, "DMC": 3.4, "ISI": 1.3, "Classes": 0 }' ``` ## File Structure ```bash . ├── application.py # Flask application ├── dependencies.py # Contains data t …