The invention provides a vehicle insurance claim rate pricing method,
system and device and a storage medium, and the method comprises the steps: obtaining and fusing a VIN code and violation data of a to-be-insured vehicle, obtaining corresponding claim rate data and claim case data from international and domestic insurance market data and same industry insurance company data, carrying out the normalization
processing, and determining the claim rate of the to-be-insured vehicle; the method comprises the following steps: obtaining corresponding
feature data, screening loss ratio data of which the
model prediction value is higher than a preset threshold value, forming a loss ratio
data set, constructing a loss ratio prediction model by using an
Adaboost algorithm fused with
logistic regression, taking loss case data as a sample, and training the prediction model in combination with the loss ratio
data set. And predicting the to-be-insured vehicle through the prediction model to obtain an insurance claim rate pricing category of the VIN code of the to-be-insured vehicle and the violation data. According to the method, a reasonable vehicle insurance pricing
system is established in combination with vehicle owner and vehicle factors, more risk factors are identified based on
generalized linear model risk modeling, and high-quality customers and low-quality customers are distinguished.