An efficient approach for road Traffic risk analysis

Abstract

Road traffic risk analysis can be broadly classified into three newlineCategories namely road traffic risk analysis on spatial data road traffic risk newlineanalysis on non spatial data road traffic risk analysis on spatial and non newlinespatial data In the first category road traffic images for different dimensions newlineare collected and analyzed to identify traffic risk spots In the second newlinecategory non spatial information such as time speed and occupancy are newlinecollected for different dimensions across the road for different time intervals newlineand these data are analyzed to identify traffic risk spots Similarly the third newlinecategory deals with both spatial and non spatial data Data mining newlinefunctionalities such as clustering classification association rule mining and newlineoutlier detections are applied to these three categories to identify the traffic newlinerisk spots The existing algorithms suffers from high dimensionality and are newlineunable to cluster accurately across the various dimensions newlineRecent progress in spatial road traffic data mining has led to the newlinedevelopment of numerous methods to mine interesting patterns and newlineknowledge from large spatial data set Clustering is the widely used data newlinemining technique and the active research area in the field of statistics pattern newlinerecognition and machine learning newline newline newline