arpitkanani/Forest-fire

Algeria Forest fire Dataset to predict FWI by Regression # 🔥 Algerian Forest Fire Weather Index Prediction ### Deployed on Render Link :fire-whether-index-predicti… ## 📌 Overview This project focuses on predicting the **Fire Weather Index (FWI)** using **regression-based machine learning models**. FWI is a continuous indicator that represents the **potential intensity and severity of forest fires**. The model is trained on the **Algerian Forest Fires dataset**, which contains meteorological and fire-related measurements from two regions of Algeria. --- ## 🎯 Problem Statement Forest fires cause significant environmental and economic damage. Accurately estimating fire severity in advance is crucial for prevention and disaster management. **Goal:** Build a machine learning regression model that predicts the **Fire Weather Index (FWI)** based on weather conditions, fire indicators, and regional information. --- ## 🌍 Dataset Description The dataset consists of daily observations collected during the **high fire-risk season (June–September)**. ### 🔹 Regions Covered - **Bejaia Region (Region = 0)** - **Sidi-Bel-Abbes Region (Region = 1)** --- ## 📊 Input Features - **Temperature** – Ambient temperature (°C) - **RH** – Relative Humidity (%) - **Ws** – Wind speed (km/h) - **Rain** – Rainfall amount (mm) - **FFMC** – Fine Fuel Moisture Code - **DMC** – Duff Moisture Code - **ISI** – Initial Spread Index - **Classes** – Fire occurrence indicator - `0` → No Fire - `1` → Fire - **Region** – Geographical region - `0` → Bejaia - `1` → Sidi-Bel-Abbes ### 🎯 Target Variable - **FWI (Fire Weather Index)** – Continuous value indicating fire severity --- ## 🧠 Machine Learning Approach - **Problem Type:** Regression - **Pipeline Steps:** - Data cleaning and preprocessing - Feature scaling - Handling regional and class indicators - Training multiple regression models - Selecting the best-performing model --- ## 📈 Model Evaluation The regression model is evaluated using: - **R² Score** - **Mean Absolute Error …