Combination of SVM-GA Method based Forest Area Estimation in Kutai National Park of East Borneo

摘要

The paper presents the combination of support vector machine (SVM) and geneticudalgorithm (GA) methods for the forest area estimation in Kutai National park of East Borneo.udThe considering variables are reboization concerning natural green, forest fire, encroachmentudand illegal logging activities. The SVM method is superior in the data classification, whileudthe GA method is well-known method in the estimation task. The input data that has beenudclassified using SVM method will be the input chromosome for the GA method. As results,udthe computational effort and convergence are fast and the method has high accuracyudestimation because only the important data is processed. The accuracy of area estimationudresult is compared with the actual data using mean absolute percentage error (MAPE).