Personal health risk score predicting method and system

Abstract

The invention provides a personal health risk score predicting method and a system. The method comprises the steps of acquiring basic medical insurance reimbursement data and corresponding participantinformation data in this year; performing data standardization on the basic medical insurance reimbursement data, the basic medical insurance detail data and the participant information data, therebyobtaining standard basic medical insurance settlement data, standard basic medical insurance settlement detail data and standard participant information data; generating four major kinds of predicting factors based on the standard basic medical insurance settlement data, the standard basic medical insurance settlement detail data and the standard participant information data, and generating a plurality of sub-predicting factors based on the four kinds of major predicting factors; performing characteristic conversion on the sub-predicting factors; performing characteristic dimension reductionon the sub-predicting factors for reducing the number of the sub-predicting factors; establishing a linear regression model based on the sub-predicting factors after characteristic selection for predicting a personal medical expense in a next year; and predicting a personal health risk score in the next year based on the predicted personal medical expense in the next year.

Description

technical field [0001] The present invention relates to the technical field of personal health risk scoring, in particular to a method and system for predicting personal health risk scoring. Background technique [0002] The basic medical insurance reimbursement data covers the personal information of the insured person, disease information, medical treatment behavior, medical expenditure, social insurance type and other fields, and the incidence of medical treatment is much higher than that of other insurances. The data granularity is fine and can Used to describe the portrait of the insured person in terms of medical health, the established personal health risk scoring model can be realized for: [0003] - Pricing underwriting of medical insurance, personal health risk score can directly reflect the risk of personal medical expenses in the next year, and accurate underwriting can be achieved for consumer medical insurance; [0004] -Auxiliary underwriting of critical illn...

AI Technical Summary

Problems solved by technology

However, in view of the limitations of data volume and data granularity, there will be insufficient prediction accuracy.

Images