@inproceedings{scholars15406,
           pages = {405--409},
           title = {Comparative analysis of support vector machine and maximum likelihood classifications using satellite images of Selangor, Malaysia},
       publisher = {Institute of Electrical and Electronics Engineers Inc.},
         journal = {2021 3rd International Sustainability and Resilience Conference: Climate Change},
             doi = {10.1109/IEEECONF53624.2021.9668109},
            year = {2021},
            note = {cited By 0; Conference of 3rd International Sustainability and Resilience Conference: Climate Change, ISRC 2021 ; Conference Date: 15 November 2021 Through 17 November 2021; Conference Code:176395},
          author = {Baig, M. F. and Ul Mustafa, M. R. and Takaijudin, H. B. and Zeshan, M. T.},
            isbn = {9781665416320},
             url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85125086974&doi=10.1109\%2fIEEECONF53624.2021.9668109&partnerID=40&md5=e06996fa8ebd42aa8570aad1e2bd7284},
        keywords = {Classification (of information); Geographic information systems; Image classification; Information management; Land use; Maximum likelihood estimation; Remote sensing; Sustainable development, Land use maps; Landuse change; Landuse classifications; LULC; Ma ximum likelihoods; Malaysia; Maximum-likelihood; Remote-sensing; Spatial classification; Support vectors machine, Support vector machines},
        abstract = {Land use maps are necessary for assessing the land use changes, that have an impact on various environmental and ecological phenomenon. Environmental processes and land use changes can now be reliably mapped and monitored due to satellite remote sensing. Land use classification is employed to assess the changes in land use patterns. Supervised classification methods are widely used because of objectivity and accuracy. The study aims to analyze the accuracy of land use maps generated using) Maximum Likelihood (ML) and Support Vector Machine (SVM) methods of land use classification. Landsat images of the state of Selangor, Malaysia from the year 2021 was used as the input dataset for the image classification. Accuracy assessment was then conducted to measure the validity of the generated classified map. The results of the classification show that SVM is more accurate than ML. The kappa coefficient obtained from SVM was 0.904, whereas for ML was 0.864. The findings of this study will offer useful information on the key aspects of land use patterns that may be applied in natural resource management and urban planning for long-term sustainability. {\^A}{\copyright} 2021 IEEE.}
}