Table 4: Prediction of forest fire risks.
Reference, Year |
Purpose |
Data |
Classifier/Predictor |
Performance |
Safi and Bouroumi [132] |
Fire weather index prediction |
Weather observations |
NN |
9% error rate |
Oliveira, et al. [135] |
Fire density |
Environmental, demographic, infra-structure, socio-economic |
RF |
96% variance explained |
Cortez and Moarias [136] |
Fire weather index prediction |
Weather observations |
SVM |
12.7% Error |
Arpaci, et al. [138] |
Fire prediction |
Weather, topology, infra-structure, socio-economic |
RF |
78% Accuracy |
Liang, et al. [139] |
Wildfire scale Prediction |
Weather and wildfire data |
LSTM |
90.9% Accuracy |
Rodrigues and Riva [140] |
Human caused wildfire occurrences |
socio-economics and economic activity, Fire causing possibilities |
LR, SVM, RF |
AUC=0.746 |
Tien Bui, et al. [141] |
Spatial Pattern of forest fires |
Weather, vegetation and infrastructure |
MARS-DFP |
86.5% Accuracy |
Qu, et al. [144] |
Fire occurance forecasting |
Weather data |
Auto-sklearn framework |
87% Accuracy |