Predicting Forest Fire Hazards Using Data Mining Techniques: Decision Tree and Neural Networks

Hamadeh, Nizar; Karouni, Ali; Daya, Bassam

doi:10.4028/www.scientific.net/amr.1051.466

Cited by 7 publications

(4 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The algorithm is based on statistical learning theory and structural risk minimization capable of addressing linear and non-linear multivariate regression problem [187]. SVM separates two opposite analogous hyperplanes and get the best separating distance between two vectors [106,127]. If the space between them is large, then the overall error of the classifier is reduced.…”

Section: Methodsmentioning

confidence: 99%

“…In some papers, different types of models have been compared in various types of geoscience application. For example SVM and random space in [78,99,103], ANN-fuzzy logic approach [48,61,81,98,104,105]; Decision tree and Neural Networks [77,92,98,105,106] have been applied elsewhere. Reports have shown that some of these approaches out-performed the conventional techniques [49,107,108].…”

Section: Research Backgroundmentioning

confidence: 99%

See 1 more Smart Citation

Data Mining and Statistical Approaches in Debris-Flow Susceptibility Modelling Using Airborne LiDAR Data

Lay

Pradhan

Yusoff

et al. 2019

Sensors

View full text Add to dashboard Cite

Cameron Highland is a popular tourist hub in the mountainous area of Peninsular Malaysia. Most communities in this area suffer frequent incidence of debris flow, especially during monsoon seasons. Despite the loss of lives and properties recorded annually from debris flow, most studies in the region concentrate on landslides and flood susceptibilities. In this study, debris-flow susceptibility prediction was carried out using two data mining techniques; Multivariate Adaptive Regression Splines (MARS) and Support Vector Regression (SVR) models. The existing inventory of debris-flow events (640 points) were selected for training 70% (448) and validation 30% (192). Twelve conditioning factors namely; elevation, plan-curvature, slope angle, total curvature, slope aspect, Stream Transport Index (STI), profile curvature, roughness index, Stream Catchment Area (SCA), Stream Power Index (SPI), Topographic Wetness Index (TWI) and Topographic Position Index (TPI) were selected from Light Detection and Ranging (LiDAR)-derived Digital Elevation Model (DEM) data. Multi-collinearity was checked using Information Factor, Cramer’s V, and Gini Index to identify the relative importance of conditioning factors. The susceptibility models were produced and categorized into five classes; not-susceptible, low, moderate, high and very-high classes. Models performances were evaluated using success and prediction rates where the area under the curve (AUC) showed a higher performance of MARS (93% and 83%) over SVR (76% and 72%). The result of this study will be important in contingency hazards and risks management plans to reduce the loss of lives and properties in the area.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Research Backgroundmentioning

confidence: 99%

Data Mining and Statistical Approaches in Debris-Flow Susceptibility Modelling Using Airborne LiDAR Data

Lay

Pradhan

Yusoff

et al. 2019

Sensors

View full text Add to dashboard Cite

show abstract

“…Two strategies of data mining has utilized with the intention of forecast and decision-making, which is Decision trees and back propagation forward neural networks [12]. Temperature, Relative Humidity, Wind Speed and Daily Precipitation are the four meteorological features being utilized.…”

Section: Introductionmentioning

confidence: 99%

An Amalgamation of Decision Tree and Cosine Kernel SYM for the Fire Detection in Forest Cover

Shenbagam¹,

Rajkumar²

2021

DSA

View full text Add to dashboard Cite

Forest fire is the crucial tragedy of the ecological system that ruins the features of ecology as well as the economy, whereas putting human lives on risk and sources the lack of biodiversity. The critical factor, to handle comparable events, is Fast detection. Another replacement was made to accomplish this task by employing local sensor-based automatic tools like the sensors presented in the meteorological stations. Meteorological factors (like rain, relative humidity, and temperature) are familiar for impacting the forest fires and various fire indexes, for instance, the Forest Fire Weather Index (FWI) that use corresponding data. It has highlighted the requirement for establishing an approach of adaptive management through current experience moreover, particular factors suggested sector-wise along with short and long term visions. This study delved into a strategy of Data Mining (DM) by integrating the Decision Tree and Cosine kernel SVM (DT-CKSVM) is proposed for forecasting the location burned by a forest fire. The testing of five individual systems of DM (include, Support Vector Machines (SVM) and Random Forests), besides four individual strategies of feature selection (spatial, temporal, components of FWI, and weather attributes) has processed, on prevailing real-world data gathered from India. Four factors of meteorology (temperature, relative humidity, rain and wind) and DT-CKSVM employed by an optimum framework, and it has potential to prophesying the region burned by small fires that are more recurrent. While analyzing the fire events through DM, it has revealed that the DM's interrelationship of trend and its interaction with environmental/meteorological factors, which brings more accurate understanding on the occurrences of the forest fire, hence simplifies in alleviation, manage, avoidance in terms of preserving our valuable forest as well as the ecosystem.

show abstract

“…For this reason, scientists worked hardly to predict forest fire in an accurate way by breaking down the non-linear relationship between fire occurrence and surrounding environment. Significant high accuracy performance results were reported using data mining tools such as: neural network [5], [6], fuzzy logic [5], and decision tree techniques [7].…”

Section: Introductionmentioning

confidence: 99%

An analytical review on the most widely used meteorological models in forest fire prediction

Hamadeh

Daya

Hilal

et al. 2015

2015 Third International Conference on Technological Advances in Electrical, Electronics and Computer Engineering (TAEECE)

Self Cite

View full text Add to dashboard Cite

At the beginning of 20 th century, scientists started to develop mathematical models in order to predict the probability of occurrence of forest fires. Meteorological parameters, such as daily temperature and humidity, were mainly used. In this paper, we review the seven most usable fire prediction indices in the world, that are Angstrom, Keetch-Byram, Modified Keetch-Byram, Canadian fire weather index, Nesterov, Modified Netserov, and Baumgartner Index. A comparative study including the mathematical equations, properties, characteristics, performanceand field of application of each model is presented. The different developed models were optimized to the local characteristics of the place of study. The problematic of suitability and compliance of indices in other regions with different conditions is discussed. Recent initiatives are finally presented. Keywords-forest fire prediction; fire weather indices; limitations and transferabilityI.

show abstract

Predicting Forest Fire Hazards Using Data Mining Techniques: Decision Tree and Neural Networks

Cited by 7 publications

References 4 publications

Data Mining and Statistical Approaches in Debris-Flow Susceptibility Modelling Using Airborne LiDAR Data

Data Mining and Statistical Approaches in Debris-Flow Susceptibility Modelling Using Airborne LiDAR Data

An Amalgamation of Decision Tree and Cosine Kernel SYM for the Fire Detection in Forest Cover

An analytical review on the most widely used meteorological models in forest fire prediction

Contact Info

Product

Resources

About