Research on road traffic safety accident data mining based on multi-objective decision-making

As an important subject in the field of traffic safety, road traffic accidents are a social and economic problem of universal concern all over the world. In order to effectively improve the accuracy and efficiency of traffic safety accident data mining, a road traffic safety accident data mining method based on multi-objective decision-making is proposed. Firstly, the relationship between traffic accident flow, vehicle speed and vehicle density in urban traffic flow data is determined according to the traffic operation; Secondly, the Pareto optimal solution set is solved according to the queuing method, and the criterion function corresponding to the VSL decision model is obtained; Then, all frequent itemsets not lower than the threshold of set support are obtained by multi-objective decision-making method, and the highest frequent itemset generated by strong association rules is determined according to the principle of multi-objective decision-making; Finally, the normal association rules and abnormal association rules of road traffic accidents are obtained and stored in the association rule base to realize road traffic safety accident data mining. The experimental results show that the accuracy of traffic safety accident data mining is 97.28%, the accuracy of accident data mining is 95.28%, and the data mining time is only 3.8s, which shows that this method has a good effect on road traffic safety accident data mining.