Shrimant05/Algerian_Forest_Fire_Prediction

# Algerian Forest Fire Prediction A Flask web app that predicts the Fire Weather Index (FWI) using a ridge regression model trained on the Algerian forest fires dataset. It includes a landing page, a full prediction form, and a Jupyter notebook documenting data preparation and model training. ## Features - Ridge regression model with standardized inputs - Clean landing page and a dedicated prediction form - End-to-end notebook for data cleaning, EDA, and training ## Project Structure - application.py: Flask app with prediction route - templates/index.html: Landing page - templates/home.html: Prediction form + results - models/: Serialized model and scaler - SLR_PROJECT.ipynb: Data prep, EDA, and modeling - requirements.txt: Python dependencies ## Data The dataset is the Algerian forest fires dataset with 244 observations across two regions (Bejaia and Sidi Bel Abbes). The target variable is FWI (continuous). Inputs include weather and fire indices such as Temperature, RH, Ws, Rain, FFMC, DMC, ISI, plus Classes and Region. ## Setup 1) Create and activate a virtual environment. 2) Install dependencies: ```bash pip install -r requirements.txt ``` ## Run the App ```bash python application.py ``` Open your browser at: - `127.0.0.1` for the landing page - `127.0.0.1` for the full prediction form ## Prediction Inputs The form expects these inputs (all numeric): - `Temperature` - `RH` - `Ws` - `Rain` - `FFMC` - `DMC` - `ISI` - `Classes` (0 or 1) - `Region` (0 or 1) ## Model The model is a ridge regression trained on cleaned data with feature scaling via `StandardScaler`. The serialized model and scaler are stored in models/. ## Notebook The notebook SLR_PROJECT.ipynb walks through: - Data cleaning and preprocessing - EDA and correlation analysis - Feature scaling - Linear, Lasso, Ridge, and ElasticNet experiments - Saving the trained ridge model and scaler ## Notes …