stefanluncanu24/Forest-Fire-Prediction-using-Decision-Trees

The GitHub project utilizes decision trees to predict forest fires using the Algerian Forest Fires Dataset obtained from the UCI Machine Learning Repository. # Forest-Fire-Prediction-using-Decision-Trees The GitHub project utilizes decision trees to predict forest fires using the Algerian Forest Fires Dataset obtained from the UCI Machine Learning Repository. The model's high accuracy and perfect recall indicate robust performance in predicting forest fires. Source ## Dataset Information The project utilizes the Algerian Forest Fires dataset from the UCI Machine Learning Repository. This dataset includes 244 instances encompassing data from two regions of Algeria, namely the Bejaia region in the northeast and the Sidi Bel-abbes region in the northwest, covering the period from June to September 2012. Detailed descriptions of the features can be found in the original research paper: - Abid, Faroudja & Izeboudjen, Nouma. (2020). Predicting Forest Fire in Algeria Using Data Mining Techniques: Case Study of the Decision Tree Algorithm. DOI: 10.1007/978-3-030-36674-2_37. - Dataset Link ## File Explanations - forest.csv - dataset used for training and testing - forest.ipynb - the code ## Libraries Required - Pandas - NumPy - Seaborn - Matplotlib - scikit-learn ## Data Preprocessing - Data is loaded and cleaned of null values. - The 'Classes' feature is encoded into binary labels: 'fire' as 1 and 'not fire' as 0. - Irrelevant features such as 'year', 'DMC', 'DC', 'FWI', and 'BUI' are dropped based on correlation analysis. ## Exploratory Data Analysis - Correlation heatmap is generated to understand the relationships between features. ## Model Building - A DecisionTreeClassifier is employed. - GridSearchCV is used to tune hyperparameters and find the best model settings based on accuracy. - The model is trained using the optimized parameters. ## Model Evaluation - The model's performance is assessed using various metrics: accuracy, recall, precision, F1-score, and confusion matrix. - **Performance Metrics:** - **Accuracy:** 0.9672131147540983 - **Confusion Matrix:** ``` [[26 2] [ 0 33]] ``` - **Recall:** 1.0 - **Pr …