Sravanisekhar/algerian-forest-fires-regression

Algerian Forest Fires Analysis using Regression Techniques Analysis of forest fire data from Bejaia and Sidi Bel-abbesusing Linear, Lasso, Ridge, and ElasticNet regression. # Algerian Forest Fires Analysis ## 📌 Overview This project analyzes the Algerian Forest Fires dataset, which includes 244 instances from two regions in Algeria: Bejaia and Sidi Bel-abbes. The data spans from June to September 2012 and contains 11 input attributes and 1 output attribute (fire occurrence). The dataset is classified into two categories: - **Fire**: 138 instances - **Not Fire**: 106 instances ## 🎯 Objectives - Preprocess the dataset for regression analysis - Apply Linear Regression, Lasso, Ridge, and ElasticNet models - Evaluate model performance and interpret results ## 📊 Dataset Information - **Total Instances**: 244 - **Regions**: Bejaia (122), Sidi Bel-abbes (122) - **Time Period**: June 2012 to September 2012 - **Attributes**: 11 input features + 1 output class (fire/not fire) ## ⚙️ Preprocessing Steps - Handling missing values - Encoding categorical variables (if any) - Feature scaling (Standardization) - Train-test split ## 🧠 Regression Models Applied - **Linear Regression**: Baseline model for prediction - **Lasso Regression**: Regularization to reduce overfitting and perform feature selection - **Ridge Regression**: Regularization to handle multicollinearity - **ElasticNet Regression**: Combines Lasso and Ridge for balanced regularization ## 📈 Evaluation Metrics - Mean Squared Error (MSE) - R-squared Score - Cross-validation scores ## 🚀 How to Run 1. Clone the repository: ```bash git clone github.com cd algerian-forest-fires-analysis