ABSTRACT

Among the yearly frequent incidences, road accidents, stands out as a major reason behind mortality across India, as a consequence of increased inhabitants and rise in the population of vehicles in our society. In our study, we made an attempt to analyse the strongly affected accident areas in the south western state, Kerala and hence discover the key factors responsible for it using the data mining techniques. In this study, the data were initially tested for heterogeneity using a two-step cluster analysis to evaluate the accident severity from 2007-2017. Later, correlation data was subsequently reduced to a smaller number of factors by applying the principal component analysis, thus identifying the major influencing factors of road crashes with massive causalities. Taking deaths and injuries into consideration, the cluster analysis clearly explained that the districts of South Kerala and Central Kerala were more affected than those of North Kerala. Principal component analysis was carried out for the entire dataset, due to the violation of sample adequacy by each cluster. Results of the analysis made it evident that, exceeding speed of the vehicles have high influence on the highway road crashes. Thus, the multivariate techniques adopted served useful in classifying high and comparatively low intensity areas and predicting the contributing factors that need to be focused to reduce accident impact on population health and in building of sophisticated public administration. Keywords: Data mining, principal component analysis, road accidents, two-step cluster analysis
Publication date: 01/09/2022

https://ijbpas.com/pdf/2022/September/MS_IJBPAS_2022_6263.pdf

Download PDF
https://doi.org/10.31032/IJBPAS/2022/11.9.6263