Modeling Forest Fires Risk using Spatial Decision Tree

Abstract

Forest firts ban long been annual eve nts in many parts or Sumatra Indonesia during the dry season. Riau Province is one of the regions in Sumatra where forl'SI fires seriously occur every year mostly because of human factors both on purpous and atddently. Forest lire models have been denlopcd for certain area using the wcightagc and criterion of ' 'ariables that involve the subjectin and qualitalin judging for variables. Determining the weights for each criterion is based on expert knowledge or tile previous ex perienced or the developers that may result too subjective models. In additlon, criteria tv1lu1tion and wei&hting method are most applied to evaluate the small problem containing few cr iteria. This paper presents our initial work in developing a spatial decision tree using the spatlal IDJ algorithm and Spatial J oin lnde.x applied in the SCART (Sp1tial Classification and Regression Trees) algorithm. The algorithm Is applied on historic forest fires data for a district in Riau namely Rokan Hilir to develop a model for forul fires risk. The modeling forest lire risk includes va riables rellltetl 10 physi c1I as well as social and economi c. The ruuh is a spn tial decision tree containing 138 leaves wilh distance to neare.~t ri\'er as the first lest attribute.