khajuriakrishna18-maker/Algerian-Forest-Fire-Prediction-using-Linear-Regression-and-Regularization-Techniques

A machine learning project that predicts the Fire Weather Index (FWI) using the Algerian Forest Fires dataset. This notebook covers data preprocessing, feature selection through correlation analysis, feature scaling, and the implementation of multiple regression models including Linear Regression, Lasso, Ridge, and Elastic Net. Algerian Forest Fire Prediction Overview This project focuses on predicting the Fire Weather Index (FWI) using the Algerian Forest Fires dataset. The notebook demonstrates a complete machine learning workflow, including data preprocessing, exploratory analysis, feature engineering, scaling, and regression model comparison. The primary goal is to evaluate different linear regression-based algorithms and determine which model provides the most accurate predictions for wildfire risk assessment. Dataset The project uses the Algerian Forest Fires Dataset, which contains meteorological and environmental attributes related to forest fire occurrences. Features Temperature Relative Humidity (RH) Wind Speed (Ws) Rain FFMC DMC DC ISI BUI Classes (Fire / Not Fire) Target Variable FWI (Fire Weather Index) Project Workflow 1. Data Preprocessing Load dataset using Pandas Remove unnecessary date columns Convert categorical fire classes into numerical values 2. Feature Engineering Split data into training and testing sets Analyze feature correlations Remove highly correlated features to reduce multicollinearity 3. Feature Scaling Apply StandardScaler to standardize feature values 4. Model Training The following regression models are implemented: Linear Regression Lasso Regression LassoCV Ridge Regression RidgeCV Elastic Net ElasticNetCV 5. Model Evaluation Models are evaluated using: Mean Absolute Error (MAE) R² Score Actual vs Predicted scatter plots Technologies Used Python Pandas NumPy Matplotlib Seaborn Scikit-Learn Jupyter Notebook Results The notebook compares multiple regularized regression techniques and evaluates their effectiveness in predicting the Fire Weather Index (FWI). Cross-validation methods such as RidgeCV, LassoCV, and ElasticNetCV are used to optimize model performance. Learning Outcomes Data preprocessing and cleaning Correlation-based feature selection Feature scaling using StandardScaler Regularization techniques (L1, L2, Elastic Net) Regression mo …