Optimizing Accounting Information Systems With Hybrid Capsule Network and Honey Badger Particle Swarm Optimization

Accounting procedures have historically been the fundamental framework for managing data environments. In this study, we aim to enhance the Accounting Information System (AIS) efficiency through optimization techniques. Our primary objective is to propose and evaluate hybrid optimization methods to improve the performance of AIS. Furthermore, the Hybrid Capsule Network with XGBoost (CNGB) model and the Hybrid Honey Badger Particle Swarm Optimization (HBPSO) method are used for search optimization while utilizing the capabilities of the PSO optimization algorithm. The data pre-processing is performed and extensive data analysis is conducted for auditing the risk assessment with predictive models. Binary classification is performed in our proposed model while utilizing the properties of the capsule network and XGBoost. Furthermore, parallel search optimization and parameter tuning are accomplished with HBPSO, which ultimately helps in efficient and reliable search results. We also conduct some experiments to evaluate the performance of our proposed models. The experimental results show that our proposed hybrid optimization model is efficient and outperforms all benchmark schemes. Our proposed model HBPSO achieves an accuracy of 95%. Besides this, the logistics regression also achieves 75% accuracy while the proposed CNGB model achieves the highest accuracy of 97.9% for auditing the information system data, which shows the effectiveness and reliability of our proposed optimization algorithm.