Assessment of data mining, multi-criteria decision making and fuzzy-computing techniques for spatial flood susceptibility mapping: a comparative study. Issue 26 (13th December 2022)

• Record Type: Journal Article
• Authors: Balogun, Abdul-Lateef; Sheng, Tan Yong; Sallehuddin, Muhammad Helmy; Aina, Yusuf A.; Dano, Umar Lawal; Pradhan, Biswajeet; Yekeen, Shamsudeen; Tella, Abdulwaheed
• Is Part Of: Geocarto international. Volume 37:Issue 26(2023)
• Page Start: 12989
• Page End: 13015
• Publication Date: 2022-12-13
• Subjects: Multi-criteria decision making -- GIS -- flood hazard -- remote sensing -- flood susceptibility -- fuzzy computing -- machine learning; Remote sensing -- Periodicals; Geographic information systems -- Periodicals; Geology -- Periodicals; Cartography -- Periodicals; 621.3678
• DOI: 10.1080/10106049.2022.2076910
• Languages: English
• ISSNs: 1010-6049

Abstract

Abstract: This study develops an Adaboost-GIS model for flood susceptibility mapping and evaluates its relative performance by undertaking a comparative assessment of the machine learning model with Multi-Criteria Decision Making (MCDM) and soft computing models integrated with GIS. An Analytic Hierarchy Process (AHP), Analytic Network Process (ANP), Fuzzy-AHP, Fuzzy-ANP and AdaBoost machine learning models were developed and integrated with GIS to classify the susceptibility of the study area. Out of 70 sample validation locations, Adaboost's performance was the best with a 95.72% similarity match with very high and high susceptibility locations followed by F-ANP, ANP, F-AHP and AHP with 95.65%, 92.75%, 81.42% and 77.14% similarity matches, respectively. It also had the highest AUC (0.864). Thus, the Adaboost machine learning, Fuzzy computing and conventional MCDM models can be adopted by stakeholders for accurately assessing flood susceptibility, thereby fostering safe and resilient cities.