Abstract :

Insurance Fraud Has Emerged As A Significant Challenge For The Global Insurance Industry, Resulting In Substantial Financial Losses And Operational Inefficiencies. Auto Insurance Fraud, In Particular, Involves Deceptive Practices Such As Exaggerated Claims, Staged Accidents, And False Reporting, Making It Difficult To Detect Using Traditional Rule-based Systems. With The Increasing Complexity Of Fraud Patterns, There Is A Growing Need For Intelligent And Adaptive Systems Capable Of Identifying Fraudulent Activities With High Accuracy.This Research Proposes An Ensemble-based Machine Learning Framework For Detecting Auto Insurance Fraud, Enhanced By An Advanced Hyperparameter Optimization Technique Referred To As BQANA (Bayesian Quantum-inspired Adaptive Neural Algorithm). The Proposed System Integrates Multiple Classification Models To Improve Predictive Performance While Leveraging BQANA For Optimal Parameter Tuning.The System Utilizes Structured Insurance Claim Data, Including Features Such As Customer History, Policy Details, Financial Attributes, And Incident Characteristics. Key Input Variables Include Months As Customer, Age, Policy Deductible, Umbrella Limit, Capital Gains, Capital Loss, Incident Severity, Number Of Vehicles Involved, Bodily Injuries, Witnesses, And Total Claim Amount. These Features Are Processed And Fed Into An Ensemble Model That Combines The Strengths Of Multiple Machine Learning Algorithms.The BQANA Optimization Technique Is Employed To Fine-tune Hyperparameters Of The Ensemble Model. Unlike Traditional Grid Search Or Random Search Methods, BQANA Dynamically Explores The Parameter Space Using Adaptive Learning Strategies, Leading To Improved Model Performance And Reduced Computational Cost. This Approach Enhances The Model’s Ability To Capture Complex Fraud Patterns And Reduces Overfitting.The System Is Implemented As A Web-based Application Using The Django Framework, Providing A User-friendly Interface For Insurance Analysts. Users Can Input Claim Details And Receive Real-time Predictions, Including The Probability Of Fraud And Classification Results. The System Also Provides Risk Indicators To Assist Decision-making.Experimental Results Demonstrate That The Proposed Ensemble Model With BQANA Optimization Achieves Higher Accuracy, Precision, And Recall Compared To Individual Models And Traditional Tuning Methods. The System Effectively Identifies Fraudulent Claims While Minimizing False Positives, Thereby Improving Operational Efficiency.The Proposed Framework Offers A Scalable And Efficient Solution For Insurance Fraud Detection. It Reduces Financial Losses, Enhances Customer Trust, And Supports Data-driven Decision-making. Furthermore, The System Can Be Extended To Other Domains Such As Healthcare And Banking Fraud Detection. In Conclusion, This Research Highlights The Effectiveness Of Combining Ensemble Learning With Advanced Hyperparameter Optimization Techniques. The Proposed System Provides A Robust And Intelligent Solution For Detecting Auto Insurance Fraud In Modern Insurance Systems.

Published:

07-4-2026

Issue:

Vol. 26 No. 4 (2026)

Page Nos:

1437-1445

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite