Comparative study of landslides susceptibility mapping methods: Multi-Criteria Decision Making (MCDM) and Artificial Neural Network (ANN)

3rd International Conference on Research Methodology for Built Environment and Engineering 2017

生态环境科学

Salleh, S.A.^1 ; Abd Rahman, A.S.A.^1 ; Othman, A.N.^1 ; Wan Mohd, W.M.N.^1

Appl. Remote Sensing and Geospatial Res. Group Centre of Studies for Surveying Science and Geomatics, Faculty of Architecture, Planning and Surveying, Universiti Teknologi MARA, Selangor, Shah Alam

40400, Malaysia^1

关键词: Accuracy assessment;    Analysis techniques;    Comparative studies;    Landslide hazard;    Local authorities;    Multi-criteria decision making;    Reliable results;    Susceptibility mapping;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/117/1/012035/pdf
DOI  :  10.1088/1755-1315/117/1/012035

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

As different approach produces different results, it is crucial to determine the methods that are accurate in order to perform analysis towards the event. This research aim is to compare the Rank Reciprocal (MCDM) and Artificial Neural Network (ANN) analysis techniques in determining susceptible zones of landslide hazard. The study is based on data obtained from various sources such as local authority; Dewan Bandaraya Kuala Lumpur (DBKL), Jabatan Kerja Raya (JKR) and other agencies. The data were analysed and processed using Arc GIS. The results were compared by quantifying the risk ranking and area differential. It was also compared with the zonation map classified by DBKL. The results suggested that ANN method gives better accuracy compared to MCDM with 18.18% higher accuracy assessment of the MCDM approach. This indicated that ANN provides more reliable results and it is probably due to its ability to learn from the environment thus portraying realistic and accurate result.