Learning a Bayesian network model for predicting wildfire behavior

Zwirglmaier, Kilian; Papakosta, Panagiota; Straub, Daniel

doi:10.1201/b16387-451

Cited by 9 publications

(8 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Overall, the best-performing method was MLP, which achieved a 65% success rate using humidity and wind speed as predictors. Zwirglmaier et al (2013) used a BN to predict area-burned classes using historical fire data, fire weather data, fire behavior indices, land cover, and topographic data. Shidik and Mustofa (2014) used a hybrid model (fuzzy C-means and back-propagation ANN) to estimate fire size classes using data from Cortez and Morais (2007); the hybrid model performed best, with an accuracy of 97.50% when compared with naïve Bayes (55.5%), DT (86.5%), RF (73.1%), KNN (85.5%), and SVM (90.3%).…”

Section: Burned-area and Fire-severity Predictionmentioning

confidence: 99%

A review of machine learning applications in wildfire science and management

et al. 2020

View full text Add to dashboard Cite

Artiﬁcial intelligence has been applied in wildﬁre science and management since the 1990s, with early applications including neural networks and expert systems. Since then the ﬁeld has rapidly progressed congruently with the wide adoption of machine learning (ML) methods in the environmental sciences. Here, we present a scoping review of ML applications in wildﬁre science and management. Our overall objective is to improve awareness of ML methods among wildﬁre researchers and managers, as well as illustrate the diverse and challenging range of problems in wildﬁre science available to ML data scientists. To that end, we ﬁrst present an overview of popular ML approaches used in wildﬁre science to date, and then review the use of ML in wildﬁre science as broadly categorized into six problem domains, including: 1) fuels characterization, ﬁre detection, and mapping; 2) ﬁre weather and climate change; 3) ﬁre occurrence, susceptibility, and risk; 4) ﬁre behavior prediction; 5) ﬁre eﬀects; and 6) ﬁre management. Furthermore, we discuss the advantages and limitations of various ML approaches relating to data size, computational requirements, generalizability, and interpretability, as well as identify opportunities for future advances in the science and management of wildﬁres within a data science context. In total, we identfied 300 relevant publications up to the end of 2019, where the most frequently used ML methods across problem domains included random forests, MaxEnt, artiﬁcial neural networks, decision trees, support vector machines, and genetic algorithms. As such, there exists opportunities to apply more current ML methods — including deep learning and agent based learning — in the wildﬁre sciences, especially in instances involving very large multivariate datasets. We must recognize, however, that despite the ability of ML methods to learn on their own, expertise in wildﬁre science is necessary to ensure realistic modelling of ﬁre processes across multiple scales, while the complexity of some ML methods, such as deep learning, requires a dedicated and sophisticated knowledge of their application. Finally, we stress that the wildﬁre research and management communities play an active role in providing relevant, high quality, and freely available wildﬁre data for use by practitioners of ML methods.

show abstract

Section: Burned-area and Fire-severity Predictionmentioning

confidence: 99%

A review of machine learning applications in wildfire science and management

et al. 2020

View full text Add to dashboard Cite

show abstract

“…The CPTs of the explanatory variables are learnt from data of the test bed area. Fire behavior model The fire behavior model presented in Figure 2 is adapted from (Zwirglmaier et al 2013). The variables are chosen from a wider range of potential variables to represent the processes influencing the resulting burnt area of a fire.…”

Section: Fire Occurrence Modelmentioning

confidence: 99%

Estimating daily fire risk in the mesoscale by means of a Bayesian network model and a coupled GIS

Papakosta¹,

Scherb²,

Straub³

et al. 2014

Advances in Forest Fire Research

View full text Add to dashboard Cite

A navegação consulta e descarregamento dos títulos inseridos nas Bibliotecas Digitais UC Digitalis, UC Pombalina e UC Impactum, pressupõem a aceitação plena e sem reservas dos Termos e Condições de Uso destas Bibliotecas Digitais, disponíveis em https://digitalis.uc.pt/pt-pt/termos. Conforme exposto nos referidos Termos e Condições de Uso, o descarregamento de títulos de acesso restrito requer uma licença válida de autorização devendo o utilizador aceder ao(s) documento(s) a partir de um endereço de IP da instituição detentora da supramencionada licença. Ao utilizador é apenas permitido o descarregamento para uso pessoal, pelo que o emprego do(s) título(s) descarregado(s) para outro fim, designadamente comercial, carece de autorização do respetivo autor ou editor da obra. Na medida em que todas as obras da UC Digitalis se encontram protegidas pelo Código do Direito de Autor e Direitos Conexos e demais legislação aplicável, toda a cópia, parcial ou total, deste documento, nos casos em que é legalmente admitida, deverá conter ou fazer-se acompanhar por este aviso. Estimating daily fire risk in the mesoscale by means of a Bayesian network model and a coupled GIS

show abstract

“…Various researchers (Li, Wang, Leung, & Jiang, 2010;Liang, Zhuang, Jiang, Pann, & Ren, 2012;Peng & Zhang, 2012a;2012b;Viglione, Merz, Salinas, & Blöschl, 2013;Vogel et al, 2013) present studies to develop flood models using BN. Although BN has to be highlighted as a powerful method to find dependencies, the challenge begins when dealing with the continuous variables (Nielsen & Jensen, 2009;Uusitalo, 2007;Zwirglmaier, Papakosta, & Straub, 2013). Dougherty, Kohavi, and Sahami (1995), Friedman and Goldsmith (1996), Aguilera, Fernández, Fernández, Rumí, and Salmerón (2011), and Vogel (2014) suggested the use of discretization to overcome this problem.…”

Section: Introductionmentioning

confidence: 99%

Fuzzy Discretization Technique for Bayesian Flood Disaster Model

Ahmad-Azami

Yusoff

Ku-Mahamud

2018

JICT

View full text Add to dashboard Cite

The use of Bayesian Networks in the domain of disaster management has proven its efficiency in developing the disaster model and has been widely used to represent the logical relationships between variables. Prior to modelling the correlation between the flood factors, it was necessary to discretize the continuous data due to the weakness of the Bayesian Network to handle such variables. Therefore, this paper aimed to propose a data discretization technique and compare the existing discretization techniques to produce a spatial correlation model. In particular, the main contribution of this paper was to propose a fuzzy discretization method for the Bayesian-based flood model. The performance of the model is based on precision, recall, F-measure, and the receiver operating characteristic area. The experimental results demonstrated that the fuzzy discretization method provided the best measurements for the correlation model. Consequently, the proposed fuzzy discretization technique facilitated the data input for the flood model and was able to help the researchers in developing effective early warning systems in

show abstract

Learning a Bayesian network model for predicting wildfire behavior

Cited by 9 publications

References 12 publications

A review of machine learning applications in wildfire science and management

A review of machine learning applications in wildfire science and management

Estimating daily fire risk in the mesoscale by means of a Bayesian network model and a coupled GIS

Fuzzy Discretization Technique for Bayesian Flood Disaster Model

Contact Info

Product

Resources

About