Soft Set Parametric-based Data Clustering for Building Data Set

Abstract

Identifying buildings for safety purposes is critical to anticipate unforeseen scenarios during a disaster. Rapid Visual Screening (RVS) is one of the procedures that can be used to determine a building's hazardous structure. The growing number of buildings necessitates grouping to provide recommendations for improving the analysis or conducting a more extensive review of the same building group. This article investigates the application of fuzzy clustering to the RVS dataset. Numerous strategies are compared, including fuzzy centroid clustering, fuzzy K-partition clustering, and multi-soft set clustering. The technique is applied to the RVS data set from Kulon Progo, Yogyakarta, which has 144 cases for grouping construction. Four clusters are formed from four distinct variables with fewer conditions: Plan Drawing, Floor Plan, Connection, and Stance. The experiment is based on the rank index, the Dunn index, and response time. The results indicate that multi-soft set-based clustering outperforms other baseline approaches. The investigator or government can utilize this information to suggest treating each cluster's 'less' variable.