tamannada26/Algerian-Forest-Fire-Prediction

I processed and cleaned the Algerian forest fire dataset, analyzing the Fire Weather Index (FWI). Using Logistic Regression, I applied cross-validation and hyperparameter tuning to assess performance. After developing the model, I saved it as pickle files and tested it on unseen data for validation. # Algerian-Forest-Fire-Prediction This project focuses on predicting forest fires in Algeria using machine learning techniques, specifically analyzing the Fire Weather Index (FWI) dataset. The primary objective is to build a predictive model that can assess the likelihood of forest fires based on various weather conditions.sing Logistic Regression, I applied cross-validation and hyperparameter tuning to assess performance. After developing the model, I saved it as pickle files and tested it on unseen data for validation. # Dataset The dataset used in this project contains data related to forest fires in Algeria, specifically from two regions: Bejaia and Sidi-Bel Abbes. It includes several weather-related variables that influence fire occurrences, such as: 1. Temperature (°C) 2. Relative Humidity (%) 3. Wind Speed (km/h) 4. Rain (mm) 5. Fine Fuel Moisture Code (FFMC) 6. Duff Moisture Code (DMC) 7. Drought Code (DC) 8. Initial Spread Index (ISI) 9. Fire Weather Index (FWI) 10. Classes: Fire/No Fire # Project Workflow **1. Data Preprocessing** **Cleaning:** - Handling missing values, outliers, and formatting issues. - Exploratory Data Analysis (EDA): Generated visualizations to gain insights into the dataset's features and distribution. - Feature Engineering: Normalizing and transforming features for better model performance. **2. Modeling** - The project uses Logistic Regression to classify whether a fire will occur based on the provided weather conditions. - Cross-validation and Hyperparameter Tuning were performed to evaluate and optimize model performance. **3. Evaluation** - The model was evaluated using standard metrics like accuracy, precision, recall, and F1-score. - Cross-validation was employed to validate the model’s robustness. - Results were compared across different hyperparameter settings to determine the best configuration. **4. Deployment** - The trained model was saved as a pickle file for future use. - It was tested on unseen data to ensure g …