Performance Evaluation on Comparison of two Classification methods of Machine Learning Techniques to extract DAI of LCLU for Disaster Management Applications

Abstract

ABSTRACT---Evaluation  is vital for the comparison of performance  analysis of the proposed Machine learning techniques such as Support Vector Machine (SVM) and Maximum Likelihood Classifiers (MLC)  are of two special classifiers that can detects Damage Assessment Index (DAI) for Land Cover Land Use (LCLU).  Error minimization is very important for the purpose of assessing the accuracy for the damage affected areas  of LCLU, minimizing error is very important aspect. The reliability of DAI for LCLU is resulting from data sensed remotely and it is depending on the precise classification. In this paper, we proposed two classifiers for disaster management applications are SVM and MLC to detect DAI.  Damage affected remote sensing data which is used as training sample datasets in proposed techniques of SVM and MLC used for each classification method. The performance of evaluation of the proposed two classification techniques indicates that SVM was more accurate than MLC. Based on the results it is confirmed that employing SVM technique is most suitable than MLC for Land Cover applications.