Data Mining Techniques on Traffic Violations
Loading...
Authors
Thapa, Santosh
Lee, Jeongkyu
Issue Date
2016-04-01
Type
Presentation
Language
en_US
Keywords
Data mining , Traffic
Alternative Title
Abstract
This paper describes use of Data mining techniques used to model traffic accidents detection. It is done by determing the blackspots by using Association Rule Mining and Clusterization algorithm. It helps to ascertain the traffic violation patterns and blackspot of traffic violations. We looked into K-means clustering with some enhancements to aid in the process of identification of patterns and blackspots. We applied these techniques to real traffic data extracted from the Montgomery County of Maryland and validated our results. We also developed a prioritized scheme for attributes here to deal with the limitations of various out of the box clustering methods and ways. This easy to implement data mining framework works with the geo-spatial plot of blackspots and helps to improve the road accidents zones.