Abstract: Claims processing is the series of steps that insurance companies follow to review, verify, and resolve claims from policyholders. An insurance company handling commercial claims has observed a rise in fraudulent cases involving various types of claims over the past few years. The government is collaborating with several organizations to detect and prevent fraudulent activity, as the amount of money that can be obtained through fraud is substantial and could cause serious problems. False accident claims can be filed to commit insurance claims fraud, a common and extensively publicized form of fraud. It happens in every category of insurance claim that is extremely serious. In our proposed work, we implement a method that uses insurance claim datasets to spot fraud and exaggerated claim amounts. Whale optimization method, also known as a hybrid Grey Wolf -based on Opti_DNN. The Grey Wolf Optimizer, was used in the construction of the hybrid algorithm. This research accomplishing the Whale Optimizer Algorithm spiral equation. To increase prediction accuracy, this paper describes how to create a Whale Optimizer and a Grey Wolf Optimizer and then combine the two for initialization, Optimization, and evaluation of the pertinent Deep Neural Network (DNN) and Particle Swarm Optimization (PSO) insurance dataset. The proposed work is then validated by comparing its Effectiveness with that of similar existing machine learning algorithms. It is evident from the comparison study that our proposed model delivers better performance in terms of various evaluation metrics such as accuracy, and false alarm rate.

Keywords: Medical Claims, Hybrid Whale and Grey Wolf Optimizer, Deep Neural Network, Particle Swarm Optimization, Fraud Detection, Claim Prediction

Pages: 742-754