RISK SEGMENTATION OF INSURANCE DATA USING MACHINE LEARNING

An entire insurance application process is very lengthy and usually it takes an insurance company about 15-30 days to process an insurance application. They have a lot of details and formalities that they have to go through like existing ailments of the customer, the salary, family history, beneficiaries, etc. So, to speed up this process a segmentation of the customer data based on risk for an insurance company is done. Where risk is defined as how likely a person is to claim his insurance. We have applied various preprocessing techniques, performed oversampling to the dataset and applied Machine learning algorithms to get the best accuracy for the dataset compared to the existing work.