Title: A hybrid fuzzy knowledge-based system for forest fire risk forecasting

Authors: Mehdi Neshat; Masoud Tabatabi; Ebrahim Zahmati; Mohhammad Shirdel

Addresses: Department of Computer Science, College of Software Engineering, Shirvan Branch, Islamic Azad University, Shirvan, Iran ' Remote Sensing and GIS Center, Sari University of Agricultural and Natural Resources, Sari, Iran ' Department of Computer Science, College of Hardware Engineering, Shirvan Branch, Islamic Azad University, Shirvan, Iran ' Department of Computer Science, College of Informatics, Shirvan Branch, Islamic Azad University, Shirvan, Iran

Abstract: Fire is one of the most important factors destroying forest ecosystems which can result in negative economic and social consequences. Quick detection can be an effective factor in controlling this destructive phenomenon. This research was aimed at designing a hybrid fuzzy expert system in order to predict the size of forest fires effectively and accurately. The data were taken from the authentic dataset named forest fire in University of California (UCI). In fact, the proposed system is a hybrid of six fuzzy inference systems with acceptable performances according to their results. The accuracy of predicting the size of fire was 81.2%.

Keywords: fuzzy inference systems; FIS; hybrid systems; forest fires; fire risk estimation; fire intensity; modelling; fuzzy KBS; knowledge-based systems; risk forecasting; fuzzy expert systems; fire size prediction.

DOI: 10.1504/IJRIS.2016.082970

International Journal of Reasoning-based Intelligent Systems, 2016 Vol.8 No.3/4, pp.132 - 154

Received: 28 Jul 2015
Accepted: 16 May 2016
Published online: 17 Mar 2017 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article