Machine Learning in Medical Insurance Prediction

Abstract

Nowadays, the trend of an ageing society is more and more obvious. Accompanied with the huge population of the elderly, the medical insurance industry has more prospects and potential. As a result, more service and business operations of insurance companies are in need. With the analysis from past data, computer algorithms help a lot in predicting the new output values, aiding data-driven business decisions, ranking of influential factors and digital computerization. Through machine learning, the insurance companies are able to make a decision flatly in premiums without having unnecessary medical expenditure. The provided models include linear regression, polynomial regression, and random forest. Through the comparation of these three models, with the output data of MAE and other indicators, we can see that polynomial regression is the best model. Within the efficient operation of this method, it can soon be prevalent among the medical industry. Avoiding problems of high cost of labor and inevitable manmade mistakes, polynomial regression aids the technical advance and statistical progress to prosper.