The Application of Binary k-Means Clustering to Identify Groups of Road Traffic Accident?s Factors in United Kingdom

摘要

Cluster analysis is a formal study of methodsand algorithms for natural grouping or clustering ofobjects according to measured or perceived intrinsiccharacteristics or similarities in each objects. The patternof the each cluster and the relationship for each clusterwere identified and then relate with the frequency ofoccurrence in the data set. This study aims to apply one ofwell-known clustering techniques, k-means clustering intobinary data set in order to cluster the factors of road trafficaccidents as the number of road accidents is increasingfrom day to day. Although there might be a list ofexpected factors that causing the road traffic accidents,none of us known which group of factors that has highestcontribution that lead to road accident. By using k-meansclustering, the patterns of road traffic accidents factorswere identified.