RIMI - Repository of the Institute for Medical Research
Institute for Medical Research
    • English
    • Српски
    • Српски (Serbia)
  • English 
    • English
    • Serbian (Cyrillic)
    • Serbian (Latin)
  • Login
View Item 
  •   RIMI
  • Institut za medicinska istraživanja
  • Radovi istraživača / Researchers' publications
  • View Item
  •   RIMI
  • Institut za medicinska istraživanja
  • Radovi istraživača / Researchers' publications
  • View Item
JavaScript is disabled for your browser. Some features of this site may not work without it.

Modeling and Mapping of Forest Fire Occurrence in the Lower Silesian Voivodeship of Poland Based on Machine Learning Methods

Thumbnail
2023
Modeling_and_Mapping_of_Forest_Fire_pub_2023.pdf (6.738Mb)
Authors
Milanović, Slobodan
Kaczmarowski, Jan
Ciesielski, Mariusz
Trailović, Zoran
Mielcarek, Miłosz
Szczygieł, Ryszard
Kwiatkowski, Mirosław
Bałazy, Radomir
Zasada, Michał
Milanović, Slađan
Article (Published version)
Metadata
Show full item record
Abstract
In recent years, forest fires have become an important issue in Central Europe. To model the probability of the occurrence of forest fires in the Lower Silesian Voivodeship of Poland, historical fire data and several types of predictors were collected or generated, including topographic, vegetation, climatic, and anthropogenic features. The main objectives of this study were to determine the importance of the predictors of forest fire occurrence and to map the probability of forest fire occurrence. The H2O driverless artificial intelligence (DAI) cloud platform was used to model forest fire probability. The gradient boosted machine (GBM) and random forest (RF) methods were applied to assess the probability of forest fire occurrence. Evaluation the importance of the variables was performed using the H2O platform permutation method. The most important variables were the presence of coniferous forest and the distance to agricultural land according to the GBM and RF methods, respectively. ...Model validation was conducted using receiver operating characteristic (ROC) analysis. The areas under the curve (AUCs) of the ROC plots from the GBM and RF models were 83.3% and 81.3%, respectively. Based on the results obtained, the GBM model can be recommended for the mapping of forest fire occurrence in the study area.

Keywords:
forest fire / ignition probability / random forest / gradient boosted machine
Source:
Forests, 2023, 14, 1, 46-
Publisher:
  • MDPI
Funding / projects:
  • Ministry of Education, Science and Technological Development, Republic of Serbia, Grant no. 200169 (University of Belgrade, Faculty of Forestry) (RS-200169)
  • Ministry of Education, Science and Technological Development, Republic of Serbia, Grant no. 200015 (University of Belgrade, Institute for Medical Research) (RS-200015)
  • Polish State Forests, grant numbers 500 477 and 500 446

DOI: 10.3390/f14010046

ISSN: 1999-4907

[ Google Scholar ]
URI
http://rimi.imi.bg.ac.rs/handle/123456789/1287
Collections
  • Radovi istraživača / Researchers' publications
Institution/Community
Institut za medicinska istraživanja
TY  - JOUR
AU  - Milanović, Slobodan
AU  - Kaczmarowski, Jan
AU  - Ciesielski, Mariusz
AU  - Trailović, Zoran
AU  - Mielcarek, Miłosz
AU  - Szczygieł, Ryszard
AU  - Kwiatkowski, Mirosław
AU  - Bałazy, Radomir
AU  - Zasada, Michał
AU  - Milanović, Slađan
PY  - 2023
UR  - http://rimi.imi.bg.ac.rs/handle/123456789/1287
AB  - In recent years, forest fires have become an important issue in Central Europe. To model the probability of the occurrence of forest fires in the Lower Silesian Voivodeship of Poland, historical fire data and several types of predictors were collected or generated, including topographic, vegetation, climatic, and anthropogenic features. The main objectives of this study were to determine the importance of the predictors of forest fire occurrence and to map the probability of forest fire occurrence. The H2O driverless artificial intelligence (DAI) cloud platform was used to model forest fire probability. The gradient boosted machine (GBM) and random forest (RF) methods were applied to assess the probability of forest fire occurrence. Evaluation the importance of the variables was performed using the H2O platform permutation method. The most important variables were the presence of coniferous forest and the distance to agricultural land according to the GBM and RF methods, respectively. Model validation was conducted using receiver operating characteristic (ROC) analysis. The areas under the curve (AUCs) of the ROC plots from the GBM and RF models were 83.3% and 81.3%, respectively. Based on the results obtained, the GBM model can be recommended for the mapping of forest fire occurrence in the study area.
PB  - MDPI
T2  - Forests
T1  - Modeling and Mapping of Forest Fire Occurrence in the Lower Silesian Voivodeship of Poland Based on Machine Learning Methods
IS  - 1
SP  - 46
VL  - 14
DO  - 10.3390/f14010046
ER  - 
@article{
author = "Milanović, Slobodan and Kaczmarowski, Jan and Ciesielski, Mariusz and Trailović, Zoran and Mielcarek, Miłosz and Szczygieł, Ryszard and Kwiatkowski, Mirosław and Bałazy, Radomir and Zasada, Michał and Milanović, Slađan",
year = "2023",
abstract = "In recent years, forest fires have become an important issue in Central Europe. To model the probability of the occurrence of forest fires in the Lower Silesian Voivodeship of Poland, historical fire data and several types of predictors were collected or generated, including topographic, vegetation, climatic, and anthropogenic features. The main objectives of this study were to determine the importance of the predictors of forest fire occurrence and to map the probability of forest fire occurrence. The H2O driverless artificial intelligence (DAI) cloud platform was used to model forest fire probability. The gradient boosted machine (GBM) and random forest (RF) methods were applied to assess the probability of forest fire occurrence. Evaluation the importance of the variables was performed using the H2O platform permutation method. The most important variables were the presence of coniferous forest and the distance to agricultural land according to the GBM and RF methods, respectively. Model validation was conducted using receiver operating characteristic (ROC) analysis. The areas under the curve (AUCs) of the ROC plots from the GBM and RF models were 83.3% and 81.3%, respectively. Based on the results obtained, the GBM model can be recommended for the mapping of forest fire occurrence in the study area.",
publisher = "MDPI",
journal = "Forests",
title = "Modeling and Mapping of Forest Fire Occurrence in the Lower Silesian Voivodeship of Poland Based on Machine Learning Methods",
number = "1",
pages = "46",
volume = "14",
doi = "10.3390/f14010046"
}
Milanović, S., Kaczmarowski, J., Ciesielski, M., Trailović, Z., Mielcarek, M., Szczygieł, R., Kwiatkowski, M., Bałazy, R., Zasada, M.,& Milanović, S.. (2023). Modeling and Mapping of Forest Fire Occurrence in the Lower Silesian Voivodeship of Poland Based on Machine Learning Methods. in Forests
MDPI., 14(1), 46.
https://doi.org/10.3390/f14010046
Milanović S, Kaczmarowski J, Ciesielski M, Trailović Z, Mielcarek M, Szczygieł R, Kwiatkowski M, Bałazy R, Zasada M, Milanović S. Modeling and Mapping of Forest Fire Occurrence in the Lower Silesian Voivodeship of Poland Based on Machine Learning Methods. in Forests. 2023;14(1):46.
doi:10.3390/f14010046 .
Milanović, Slobodan, Kaczmarowski, Jan, Ciesielski, Mariusz, Trailović, Zoran, Mielcarek, Miłosz, Szczygieł, Ryszard, Kwiatkowski, Mirosław, Bałazy, Radomir, Zasada, Michał, Milanović, Slađan, "Modeling and Mapping of Forest Fire Occurrence in the Lower Silesian Voivodeship of Poland Based on Machine Learning Methods" in Forests, 14, no. 1 (2023):46,
https://doi.org/10.3390/f14010046 . .

DSpace software copyright © 2002-2015  DuraSpace
About RIMI | Send Feedback

OpenAIRERCUB
 

 

All of DSpaceCommunitiesAuthorsTitlesSubjectsThis institutionAuthorsTitlesSubjects

Statistics

View Usage Statistics

DSpace software copyright © 2002-2015  DuraSpace
About RIMI | Send Feedback

OpenAIRERCUB