Predicting Size of Forest Fire Using Hybrid Model

This paper outlines a hybrid approach in data mining to predict the size of forest fire using meteorological and forest weather index (FWI) variables such as Fine Fuel Moisture Code (FFMC), Duff Moisture Code (DMC), Drought Code (DC), Initial Spread Index (ISI), temperature, Relative Humidity (RH), wind and rain. The hybrid model is developed with clustering and classification approaches. Fuzzy C-Means (FCM) is used to cluster the historical variables. The clustered data are then used as inputs to Back-Propagation Neural Network classification. The label dataset having value greater than zero in fire area size are clustered using FCM to produce two categorical clusters,i.e.: Light Burn, and Heavy Burn for its label. On the other hand, fire area label with value zero is clustered as No Burn Area. A Back-Propagation Neural Network (BPNN) is trained based on these data to classify the output (burn area) in three categories, No Burn Area, Light Burn and Heavy Burn. The experiment shows promising results depicting classification size of forest fire with the accuracy of confusion matrix around 97, 50 % and Cohens Kappa 0.954. This research also compares the performance of proposed model with other classification method such as SVM, Naive Bayes, DCT Tree, and K-NN that showed BPNN have best performance.

[1]  Youssef Safi,et al.  A neural network approach for predicting forest fires , 2011, 2011 International Conference on Multimedia Computing and Systems.

[2]  Imas Sukaesih Sitanggang,et al.  Hotspot occurrences classification using decision tree method: Case study in the Rokan Hilir, Riau Province, Indonesia , 2010, 2010 Eighth International Conference on ICT and Knowledge Engineering.

[3]  Song Weiguo,et al.  A study of forest fire danger prediction system in Japan , 2004 .

[4]  Ian H. Witten,et al.  Data mining: practical machine learning tools and techniques with Java implementations , 2002, SGMD.

[5]  Abdul Rahim Nik,et al.  Pattern clustering of forest fires based on meteorological variables and its classification using hybrid data mining methods , 2011 .

[6]  Maitreyee Dutta,et al.  Neural network based handwritten hindi character recognition system , 2009, COMPUTE '09.

[7]  J. C. Dunn,et al.  A Fuzzy Relative of the ISODATA Process and Its Use in Detecting Compact Well-Separated Clusters , 1973 .

[8]  Lazaros S. Iliadis,et al.  An intelligent system employing an enhanced fuzzy c-means clustering model: Application in the case of forest fires , 2010 .

[9]  Imad H. Elhajj,et al.  Artificial intelligence for forest fire prediction , 2010, 2010 IEEE/ASME International Conference on Advanced Intelligent Mechatronics.

[10]  J. Carlin,et al.  Bias, prevalence and kappa. , 1993, Journal of clinical epidemiology.

[11]  A. Eleyan,et al.  PCA and LDA Based Neural Networks for Human Face Recognition , 2007 .

[12]  Amparo Alonso-Betanzos,et al.  An intelligent system for forest fire risk prediction and fire fighting management in Galicia , 2003, Expert Syst. Appl..

[13]  Yusuf Sönmez,et al.  A data fusion framework with novel hybrid algorithm for multi-agent Decision Support System for Forest Fire , 2011, Expert Syst. Appl..

[14]  James C. Bezdek,et al.  Pattern Recognition with Fuzzy Objective Function Algorithms , 1981, Advanced Applications in Pattern Recognition.

[15]  Jiawei Han,et al.  Data Mining: Concepts and Techniques , 2000 .

[16]  Dilip Kumar Pratihar,et al.  A Comparative Study of Fuzzy C-Means Algorithm and Entropy-Based Fuzzy Clustering Algorithms , 2011, Comput. Informatics.

[17]  Ian H. Witten,et al.  Data mining: practical machine learning tools and techniques, 3rd Edition , 1999 .

[18]  P. Cortez,et al.  A data mining approach to predict forest fires using meteorological data , 2007 .

[19]  Imad H. Elhajj,et al.  Efficient forest fire occurrence prediction for developing countries using two weather parameters , 2011, Eng. Appl. Artif. Intell..