A Comparative Evaluation of Classification Methods in the Prediction of Road Traffic Accident Patterns
It is important to study the nature of the associations between road, environmental, and traffic factors and motor vehicle crashes, with the aim to understand the reasons for crashes and to better predict their occurrence, hi this research work we evaluate the performance of several kinds of decision tree algorithms viz. C4.5, ID3, REPTree, Random Tree, Decision Stump, J48 and QUEST, for predicting road traffic accident patterns. For the study the road accident training dataset of Great Britain is obtained from the STATS 19 data collection system, maintained by the government of United Kingdom (UK). The experimental results show that when various decision trees are applied in predicting the road traffic accident patterns C4.5 Tree is the best only in terms of accuracy, Decision Stump is the best only considering speed, Random Tree is the optimal choice considering both accuracy and speed. The results have been evaluated using the accuracy measures such as Recall and Precision.
Data Mining Classification Algorithms Error Rates Road Accident Data Accident Patterns Accuracy Measures Precision Recall
S. Shanthi R.Geetha Ramani
Senior Lecturer (Ph.D.Scholar), Department of Computer Science and Engineering Rajalakshmi Institute Professor and Head, Department of Computer Science and Engineering Rajalakshmi Engineering College,
国际会议
哈尔滨
英文
162-166
2012-05-19（万方平台首次上网日期，不代表论文的发表时间）