Author : Rohan yashraj gupta, phani krishna kandala, satya sai mudigonda, pallav kumar baruah
Keyword : —comprehensive fraud management, emerging experience, extreme value theory, behavioural finance, classification trees, logistic regression, suspicious scoring, spectral clustering, peak analysis, machine learning, blockchain and distributed computi
Subject : Business
Article Type : Original article (research)
Article QR Code
Article File : Full Text PDF
Abstract : Fraud acts as a major deterrent to a company’s growth if uncontrolled. It challenges the fundamental value of “Trust†in an Insurance business. This concern must be addressed on priority else, it brings down the castle of the insurance business. The regulation provides powers to authorities to act on fraud. Currently, this effort within most organizations happens discretely, involving those unconnected with actuarial or technology. In fact, actuarial techniques are powerful tools that help to bring efficiency and to target the right areas to deploy the right level of resources for fraud investigation. An effective solution approach to tackle this challenging problem is provided in this work which is empirically tested. In this work, we propose comprehensive fraud management (CFM) framework using actuarial techniques and AI Technology that helps increase fraud detection rate in comparison with other proposed models available in the literature. This framework includes three stages: • Stage 1: Automate Fraud identification using triggers specific to individual LoB and rule engine. This is a prevention stage. • Stage 2: Integrate Statistical/ Actuarial Techniques and Technology to identify fraud. Statistical/ Actuarial techniques for fraud detection include techniques such as classification trees, logistic regression, suspicious scoring, significance testing, random sampling, clustering, linear regression, peak analysis, extreme value theory etc. Technologies that are effective in detecting fraud include machine learning, deep learning, blockchain and distributed systems etc. This is a fraud detection stage. • Stage 3: Further analyse results from stage 2 to create a new set of fraud identification triggers. This adds on to the existing set of triggers in Stage 1 and increases the fraud detection rate in the subsequent runs of CFM. Proof of concept presented here on Motor line of business can be tested and extended to other lines of business or industries.
Article by : Rohan Yashraj Gupta
Article add date : 2021-01-24
How to cite : Rohan yashraj gupta, phani krishna kandala, satya sai mudigonda, pallav kumar baruah. (2021-January-24). A framework for comprehensive fraud management using actuarial techniques. retrieved from https://www.openacessjournal.com/abstract/595