Claim risk assessment method, device and server

Abstract

The embodiment of the present application provides a claim settlement risk assessment method, device, and server. By using various behavior data of users to be estimated, the initial feature data of the first quantitative dimension is obtained through preprocessing, and features are screened by feature selection algorithms. Obtain the feature variables of the second quantity dimension and divide them into feature sets of the third quantity type, so as to use machine learning algorithms to calculate the feature variables contained in the feature sets of the third quantity type to determine the target of the user to be estimated payout ratio. It can be seen that when predicting the user's compensation risk, this application fully considers the influence of various aspects of the user to be estimated on the claim risk, and obtains thousands of input variables of the prediction model, which reduces the influence of individual input variable disturbances on the prediction results. Interference improves the accuracy and stability of user compensation risk assessment, so that business personnel can designate reasonable and reliable insurance and claim settlement plans for users to be assessed.

Description

technical field [0001] This application relates to the field of insurance applications, in particular to a claim settlement risk assessment method, device and processing server. Background technique [0002] Since the reform and opening up, with the continuous development and growth of my country's insurance industry, the functions and roles of insurance in the economic society have become more and more prominent. In particular, auto insurance (referred to as auto insurance) has developed rapidly in more than 30 years, contributing to economic and social development. It provides an important guarantee for the stability of people's lives. However, in recent years, the comprehensive ratio of the industry has remained high, and the development concept of auto insurance has gradually changed from "sales is king" to "fine management". Moreover, in the current auto insurance rate models commonly used in the industry, the data dimensions used for pricing are usually relatively smal...

AI Technical Summary

Problems solved by technology

[0006] In view of this, the embodiment of the present application provides a claim settlement risk assessment method, device and server, which solves the technical problem that the prior art cannot reasonably set auto insurance premiums according to the characteristics of different policyholders

Images