vnktadithya/Algerian-forest-fire-FWI-prediction

# 🔥 Predicting Forest Fire Weather Index (FWI) using Regression An **end-to-end Machine Learning project** focused on predicting the **Forest Fire Weather Index (FWI)** for two regions of Algeria using meteorological and environmental data. This project covers the complete ML lifecycle — from **data cleaning and EDA** to **model training, evaluation, and deployment-ready inference using Flask**. --- ## 📌 Project Overview Forest fires are highly influenced by weather conditions. The **Forest Fire Weather Index (FWI)** is a numeric indicator used to estimate fire risk. In this project, multiple **regression models** are trained and evaluated to accurately predict the FWI value for two Algerian regions: - **Bejaia Region** - **Sidi-Bel Abbes Region** The best-performing model is selected, serialized, and integrated into a **Flask-based web application** with a minimal HTML frontend. --- ## 🎯 Objectives - Perform data cleaning and preprocessing on real-world fire weather data - Conduct Exploratory Data Analysis (EDA) - Apply feature scaling and feature engineering - Train and compare multiple regression models - Select the best model based on evaluation metrics - Build a simple web interface for inference using Flask --- ## 📊 Dataset - **Source:** Kaggle - **Link:** Algerian Forest Fires Datas… - **Type:** Public dataset - **Regions Covered:** Bejaia & Sidi-Bel Abbes (Algeria) The dataset contains meteorological attributes such as temperature, humidity, wind speed, rainfall, and related fire indices. --- ## 🧠 Machine Learning Pipeline The Jupyter Notebook (`ML_project_using_Regression.ipynb`) contains the complete pipeline: 1. **Data Cleaning** - Handling missing values - Correcting data types - Removing inconsistencies 2. **Exploratory Data Analysis (EDA)** - Distribution analysis - Correlation heatmaps - Feature-target relationships 3. **Feature Engineering** - Feature selection - Feature scaling us …