Bridge strain data analysis using density based clustering technique

Density based clustering technique, DBSCAN is used to extract knowledge and decision rules from large test data sets of a large steel truss bridge over river Brahmaputra. The analyses are based on variations of intra and inter cluster densities, shape of cluster distributions and total number of data points. DBSCAN parameters are optimized for obtaining a good clustering result and thereby identifying data density and noisy data elements of a data set.