Fraud Analysis
案例介绍
Fraud Analysis
案例介绍
We found a correlation between claim size and monthly premium using ggplot scatter plots. For young drivers under 27, males had a higher fraud rate than females. An interval estimate showed that many fraudulent claims had less than 8 months since the previous claim, suggesting that stricter auditing should be applied. The Weibull distribution was found suitable for modelling the months since the last claim for fraudulent cases. Bayesian analysis estimated fraud rates of about 37% for young males and 33% for young females. Therefore, enhanced screening is recommended for claims from young male drivers with less than 8 months since their previous claim.
案例图片
1小时精推人才
需求方专属客服，免费梳理匹配
