Modelagem de risco de incêndios florestais: uma visão geral

Abstract

The fire risk assessment is important to establish the priority risk zones to have a better planning of prevention and firefighting activities. So this paper has the aim to compare different approaches to modeling the fire risk. It was discussed four approaches the Multicriteria Analysis, Fuzzy Logic, Logistic Regression, Neural Networks and Monte Alegre Formula. The neural networks have a better performance, but your execution is more complex because depends in computational knowledge, and time to test the better structure for the network. So the choose of the methodology to be use depends of the aim of the work and the time to spend in the modeling.