Conclusion and suggestions
In this paper, we proposed a framework for analyzing accident patterns for different
types of accidents on the road which makes use of K modes clustering and association
rule mining algorithm. The study uses 11,574 accidents that have occurred on Dehradun
district road network during 2009 to 2014. K modes clustering find six cluster (C1–C6)
based on attributes accident type, road type, lightning on road and road feature. Association
rule mining have been applied on each cluster as well as on EDS to generate rules.
Strong rules with high lift values are taken for the analysis. Rules for every cluster reveal
the circumstances associated with the accidents within that cluster. These rules are