ashir1S/Algerian_forest_fires

This end-to-end machine learning project builds a regression pipeline to predict the Fire Weather Index (FWI) using the Algerian Forest Fires dataset. It covers data cleaning, exploratory analysis, and feature engineering, then trains and compares Linear, Ridge, Lasso, and ElasticNet models with cross-validated tuning to find the best predictor. # 🔥 Algerian Forest Fires — FWI Regression Pipeline A regression pipeline for predicting the **Fire Weather Index (FWI)** from the Algerian Forest Fires dataset, covering data cleaning, exploratory analysis, feature engineering, and a comparison of Linear, Ridge, Lasso, and ElasticNet regression models with cross-validated hyperparameter tuning. --- ## 📑 Table of Contents 1. Project Overview 2. The Big Picture — End-to-End Pipeline 3. Repository Structure 4. Step-by-Step Walkthrough 5. Model Comparison / Results 6. Key ML Concepts Covered 7. Limitations 8. Tech Stack 9. How to Run 10. Author --- ## 1. Project Overview **Dataset:** Algerian Forest Fires Dataset — 244 records from two regions of Algeria (Bejaia in the northeast, Sidi-Bel Abbes in the northwest), 122 records each, collected June–September 2012. Each record has 11 weather/fire-index attributes plus a `Classes` label (`fire` / `not fire`). **Target variable:** `FWI` (Fire Weather Index) — a continuous score, so this is framed as a **regression problem**. The dataset also includes a binary `Classes` column, which would support a classification formulation ("will there be a fire?"). This project instead predicts the continuous FWI score; a classification variant using `Classes` as the target is a natural extension. **Two notebooks, two jobs:** | Notebook | Role | |---|---| | `01_Data_Cleaning_and_EDA.ipynb` | Loads the **raw** CSV, cleans it, explores it (EDA), and writes out a cleaned CSV | | `02_Model_Training.ipynb` | Loads the **cleaned** CSV, engineers features, trains 4 regression models (+ 3 cross-validated variants), and evaluates them | --- ## 2. The Big Picture — End-to-End Pipeline The diagram below summarizes the complete workflow, from raw data to final model selection. ```mermaid flowchart TD A["📥 Raw Data Algerian_forest_fires_dataset_UPDATE.csv"] --> B["🧹 Data Cleaning (Ridge__Lasso_Regression.ipynb)"] B --> B1["Tag rows by Region Fix column dtypes Drop the blank separato …