YShukla2024/Algerian_Forest_Fire_project

# 🌲 Algerian Forest Fire Detection – End-to-End Machine Learning Project ## 📌 Project Overview This project is an **end-to-end machine learning pipeline** for predicting the **Fire Weather Index (FWI)** using the **Algerian Forest Fire Dataset**. It covers **data cleaning, exploratory data analysis (EDA), feature engineering, model training, evaluation, and deployment preparation**. The final model selected is **RidgeCV Regression**, achieving an **accuracy of ~98%**. --- ## 📊 Dataset - **Source**: Algerian Forest Fire Dataset – UCI Machine Learning Repository - **Features**: - **Spatial**: Region, Month, Day - **Meteorological**: Temperature, Relative Humidity, Wind Speed, Rain - **FWI Components**: FFMC, DMC, DC, ISI - **Target Variable**: FWI (Fire Weather Index) --- ## 🛠 Workflow 1. **Data Cleaning & Preprocessing** - Handled missing values - Standardized column names - Converted data types - Encoded categorical variables 2. **Exploratory Data Analysis (EDA)** - Distribution plots - Correlation heatmaps - Outlier detection 3. **Feature Engineering** - Scaling numerical features - Removing redundant features 4. **Model Training & Evaluation** - **Linear Regression** - **LassoCV** - **ElasticNetCV** - **RidgeCV** ✅ *(Selected – Best Performance)* 5. **Model Selection** - RidgeCV chosen due to **highest R² score (~0.98)** and stability. 6. **Deployment Preparation** - Created Flask app (`application.py`) - Saved trained model in `Model/ridge_model.pkl` - Added `requirements.txt` for reproducibility --- ## 📈 Model Performance | Model | R² Score | |----------------|---------| | Linear Regression | ~97% | | LassoCV | ~95% | | ElasticNetCV | ~94% | | **RidgeCV** | **~98%** ✅ | --- ## 🚀 How to Run Locally ### 1️⃣ Clone the repository ```bash git clone github.com /Algerian_Forest_Fire_Detection.git cd Algerian_Forest_Fire_Detection