Forecasting of forest fires using machine learning techniques: a comparative study

Abstract

Listen

Translate article

Forest fires are considered a critical issue to the environment, economy and a risk threatening people's life. Due to its rapid spread rate and brutality, controlling it prematurely is of great significance; therefore, the introduction of artificial intelligence for forecasting forest fires can be of a remarkable effect in reducing the unpleasant consequences of this catastrophe. In this article, three ML methods namely Logistic Regression (LR), Decision Tree Classifier (DTC), and Gaussian Naïve Bayes (GNB) were used to compare and evaluate their performance at forecasting forest fires. The experimental results showed that the LR outperformed the others achieving the best performance at 100% for the accuracy, F-measure, sensitivity, precision, and ROC-AUC.