Landslide Susceptibility Mapping using Association Rule Mining Based Apriori Algorithm and Multiple Clustering Algorithms

Abstract

Haphazard development activities on mountain slopes and inadequate attention to construction aspects have led to increasing landslides and sustaining damages to the lives and the infrastructure. According to the National Research Building Organization (NBRO) reports, within the study area, nearly 3275 sq.km of the area expanded over the Ratnapura District; and 2178 sq.km area is to be highly prone to landsliding. If the appropriate investigations were performed in time, most of the landslides could be predicted relatively. This study aims to discover the real extent and severity of the landslide processes and risk evaluation within the study area. Machine Learning Approach based on Association Rule Mining and multiple Clustering algorithms were combined and implemented for the final prediction. This study possesses a strong capability to predict landslides risk by considering causative factors slope, Landuse, Geology, Soil material, elevation, intensity, and triggering factor; rainfall. Apriori Association rule algorithm, K-mean and Expectation Maximization (EM) clustering algorithms are the highest-ranking prediction algorithms. While applying the EM clustering method showed accuracy over 84% of the results with high speed and time taken to build was 0.66 seconds. The K-means algorithm gained the highest accuracy over 92% was applied and time taken to build was 1.58 seconds, though it was more time-consuming than EM algorithm. While applying the Apriori algorithm to obtain the best results, therefore ten (10) efficient prediction rules have found to fulfil the ultimate goal of this research. Moreover, the results show that the EM and Apriori algorithms have the best degree to fit for the Landslide susceptibility mapping.