Barunwason/-Algerian-Forest-Fire-Prediction

# 🔥 Algerian Forest Fire Prediction This project predicts the **Fire Weather Index (FWI)** using meteorological and fire behavior indicators from the Algerian Forest dataset. A **Linear Regression model** (with Ridge and Lasso regularization) was trained after extensive feature engineering and EDA. A **web interface** built with **Flask** allows users to input conditions and receive real-time FWI predictions. --- ## 📌 Project Overview - **Objective**: Predict the Fire Weather Index (FWI) based on weather and environmental features. - **Target Variable**: FWI - **Tech Stack**: - Python (Pandas, Scikit-learn, Matplotlib, Seaborn) - Linear Regression (Ridge and Lasso) - Flask for web deployment - HTML/CSS for the frontend --- ## 🔍 Feature Engineering and EDA Before training the model, a thorough **Exploratory Data Analysis (EDA)** and **feature engineering** process was performed: 1. **Data Cleaning**: - Removed duplicate rows. - Handled missing values by imputation or deletion based on relevance. 2. **Correlation Analysis**: - Calculated the correlation matrix to identify highly correlated features. - Dropped highly correlated variables to reduce multicollinearity and improve model performance. 3. **Feature Selection**: - Retained only relevant features with good predictive power for FWI. --- ## 🤖 Model Building - **Algorithms Used**: - Ridge Regression - Lasso Regression - **Process**: - Standardized input features. - Trained models and evaluated performance using appropriate metrics. - Saved the best-performing model using `pickle` for easy deployment. --- ## 🌐 Web Interface with Flask A user-friendly **web page** allows users to input environmental parameters such as temperature, humidity, wind speed, and fire behavior indices. - The input is sent to the Flask backend. - The backend loads the **trained model from the pickle file**. - The FWI prediction is displayed on the webpage in real-time. > Example inputs: > Temperature, Relative Humidity, W …