Prediction of environmental hazards and application of risk maps to wolf (Canis lupus) attacks on human in Hamedan province

Today, modeling risk prediction as an effective strategy to protect the interests of human and wildlife populations is considered. The application of output maps such models to determinant the distribution range of conflicts; cost estimates and success of control programs are very beneficial. In the present study, we present a risk map for gray wolf (Canis lupus) attacks on human in Hamedan province. Thus, evaluation the performance of maximum entropy (MAXENT) modeling method to two methods presence/absence of artificial neural networks (ANN) and Support Vector Machine model (SVM). Based on predictions obtained north and center of the province are facing with high occurrence of wolf attacks. The results showed the important effective factors in distribution of wolf attacks are land use, human density and distance from the road. Between these factors, the land use was identified as the most important parameter. In addition, according to the results, the maximum entropy method, amount allocated area under the curve was greater (AUC =0.88) that indicates the high efficiency of this method in prediction of high risk areas in Hamedan province.