Correlation Analysis of Traffic Accident Factors based on Mean Clustering

In order to effectively improve the mining efficiency of traffic accident factors and enhance the clarity of mining results, an association analysis method based on mean clustering is proposed. Firstly, the method generalizes the accident data, extracts the main accident attributes, and uses K-means clustering to classify the accident level according to the number of casualties; Based on different accident levels, the improved Apriori algorithm is used for association analysis to mine the main contributing factors. The experiment uses the British government public data set and multiple data mining algorithms for quantitative and qualitative analysis. The results show that the mining efficiency of the combined algorithm has been significantly improved, and the correlation results can more intuitively reflect the relationship between accident factors and accident severity, which is suitable for traffic accident profile analysis.