The Identification Algorithm and Model Construction of Automobile Insurance Fraud Based on Data Mining

Abstract

Currently, insurance fraud spreads quickly in the domestic and foreign field, especially in the field of automobile insurance, so that we need more efficient and accurate technology to anti automobile insurance fraud. Therefore, this paper studied the data mining technology to anti automobile insurance fraud. The improved outlier detection method based on the nearest neighbor with pruning rules was applied to automobile insurance fraud, and the improved auto insurance fraud identification model and the corresponding algorithm were established, the association rules were used to mine the law of auto insurance fraud. Finally the method has been verified by experimental analysis. The experimental results show that the improved algorithm of automobile insurance fraud identification had the advantages of low time complexity, high recognition rate, high accuracy and low impact on the K value of the algorithm.