Trend and time series cluster analysis of crime incidences in india using dynamic time warping and hierarchical clustering method

Crime is an unavoidable social phenomenon prevailing more or less in every social structure. The incidences of crime either can be minimized or controlled, however, it is difficult to eradicate it completely. To observe crime pattern of a particular place, trend analysis plays vital role in examining shape and structure of the data. In crime study, it provides help in solving issues like crime mapping, policy-prescription, prediction, etc. Trend analysis in crime data provides comprehensive as well as extensive observations of the statistics, because of which behavioral study of the same becomes trouble-free. In this paper, an attempt has been made to analyze crime data of India for last two decades. To study the trend, Dynamic Time Warping (DTW) technique has been exercised. Hierarchical clustering of all the states and UTs has been performed as well based on crime rates to figure out places with similar pattern. Moreover, simple statistical analysis has also been carried out for better understanding of the scenario.