An evolved fuzzy logic system for fire size prediction
Alan Fowler, Ankur M. Teredesai, Martine De Cock · 2009
The accurate prediction of forest fire size is important in order to issue adequate and timely warnings and to allocate fire-fighting assets efficiently and effectively. A forest fire data set collected in Portugal has recently become available as a benchmark for experimental validation of data mining techniques to tackle this problem. In this paper, we explore the suitability of a fuzzy rule based system to solve the forest fire size prediction problem. Since we have no specific domain expertise, we evolve the fuzzy rules as well as the membership functions automatically using genetic algorithms. The results clearly demonstrate the utility of the evolved fuzzy rule based system.