ADITI-VERMA05/Algerian-Forest-Fires-Prediction

# 🌲 Algerian Forest Fires Prediction This repository contains an **end-to-end machine learning project** using the **Algerian Forest Fires dataset**. The goal is to predict whether a fire occurs based on meteorological conditions using various supervised ML models. --- ## 📌 Repository Structure - `2.0-EDA And FE Algerian Forest Fires.ipynb` → Exploratory Data Analysis (EDA) & Feature Engineering - `3.0-Model Training.ipynb` → Model training, evaluation, and saving artifacts - `Algerian_forest_fires_cleaned_dataset.csv` → Cleaned dataset after preprocessing - `Algerian_forest_fires_dataset_UPDATE.csv` → Updated dataset used for training - `ridge.pkl` → Saved Ridge Regression model - `scaler.pkl` → Saved StandardScaler object --- ## 🔎 Project Workflow 1. **Data Preprocessing & Cleaning** - Handling missing values and formatting issues - Preparing datasets for modeling 2. **Exploratory Data Analysis (EDA)** - Statistical summaries and visualization - Correlation analysis of meteorological features 3. **Feature Engineering** - Scaling numerical features - Encoding categorical variables 4. **Model Training** - Ridge Regression and other ML algorithms - Hyperparameter tuning - Saving trained models as `.pkl` 5. **Model Evaluation** - Metrics: Accuracy, Precision, Recall, F1-score - Visualizations: confusion matrix, ROC-AUC 6. **Deployment Ready Artifacts** - Exported model (`ridge.pkl`) - Exported scaler (`scaler.pkl`) --- ## ⚙️ Installation Clone the repo and install dependencies: ```bash git clone github.com cd algerian-forest-fires-prediction pip install -r requirements.txt