TabNet Interpretable Deep Structure for Adaptive Recognition of Audit Voucher Anomalies
Main Article Content
Abstract
This study proposes an interpretable deep learning framework for adaptive recognition of audit voucher anomalies based on the TabNet architecture. The framework addresses the dual challenges of decision transparency and environmental adaptability in intelligent auditing systems. Structured audit voucher data are processed through sequential attention layers to generate traceable feature importance masks and contribution paths. A sliding timewindow mechanism is introduced to accumulate historical feature importance distributions, while a dynamic threshold calibration strategy based on statistical characteristics enables adaptive anomaly boundary adjustment under concept drift conditions. The proposed framework integrates feature selection, temporal distribution analysis, and anomaly scoring to achieve transparent and adaptive anomaly recognition. Experimental results demonstrate that the detection accuracy reaches 92.3% for amount logic conflicts and 93.1% for supplier-related anomalies, while all anomaly categories achieve F1-scores above 87%. In addition, the verification time for supplier anomalies is reduced from 19.6 min to 8.4 min, resulting in a 57.1% improvement in audit verification efficiency. The average importance of transaction amount features reaches 0.883 in financial audit scenarios, confirming the model’s ability to adaptively focus on critical business characteristics. The proposed framework can be integrated with wireless data acquisition infrastructures, intelligent financial information systems, and edge-computing-enabled audit platforms to support realtime anomaly monitoring and interpretable decision-making in large-scale financial auditing environments.
Downloads
Article Details

This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
References
N. T. Popoola, “Big Data-Driven Financial Fraud Detection and Anomaly Detection Systems for Regula-Tory Compliance and Market Stability,” Int. J. Comput. Appl. Technol. Res, vol. 12, no. 9, pp. 32-46, 2023, doi: 10.7753/IJCATR1209.1004.
S. C. Friday, C. I. Lawal, D. C. Ayodeji, and A. Sobowale, “Reviewing the effectiveness of digital audit tools in enhancing corporate transparency,” International Journal of Advanced Multidisciplinary Research and Studies, vol. 6, no. 4, pp. 1679-1689, 2024, doi: 10.62225/2583049X.2024.4.6.4099.
A. Al-Omush, A. Almasarwah, and A. Al-Wreikat, “Artificial Intelligence in Financial Auditing: Redefining Accuracy and Transparency in Assurance Services,” EDPACS, vol. 70, no. 6, pp. 1-20, 2025, doi: 10.1080/07366981.2025.2459490.
Q. Chen, J. Lin, X. Chen, and X. Wang, “Research on the Application of RPA+ AI Technology in the Con-struction of Paperless Intelligent System for Finance in Colleges and Universities,” Journal of Mod-ern Social Sciences, vol. 1, no. 2, pp. 368-383, 2025, doi: 10.5281/zenodo.14553755.
S. K. Afadzinu, L. D. David, I. Pothaczky Racz, and F. Jemimah, “The impact of technological innovations on audit transparency, objectivity, and assurance in the digital era,” Journal of Infrastructure Policy and Development, vol. 8, no. 14, pp. 8241-8262, 2024, doi: 10.24294/jipd8241.
G. Chaudhary, “Unveiling the black box: Bringing algorithmic transparency to AI,” Masaryk University Journal of Law and Technology, vol. 18, no. 1, pp. 93-122, 2024, doi: 10.5817/MUJLT2024-1-4.
D. Goswami, “ADVANCING THREAT DETECTION THROUGH ARTIFICIAL INTELLI-GENCE AND MACHINE LEARNING ENHANCED CYBERSECURITY AUDITS,” Ameri-can Journal of Scholarly Research and Innovation, vol. 4, no. 1, pp. 428-457, 2025, doi: 10.63125/gb5s3f54.
D. Brand, M. E. Hoffmann, and J. Van der Merwe, “Development of effective methods and tools for the auditing AI algorithms by Supreme Audit Institutions,” JeDEM-eJournal of eDemocracy and Open Government, vol. 17, no. 3, pp. 106-140, 2025, doi: 10.29379/jedem.v17i3.1049.
H. Zhong, D. Yang, S. Shi, L. Wei, and Y. Wang, “From data to insights: the application and challenges of knowledge graphs in intelligent audit,” Journal of Cloud Computing, vol. 13, no. 1, pp. 114-140, 2024, doi: 10.1186/s13677-024-00674-0.
V. Ganapathy, “AI in auditing: A comprehensive review of applications, benefits and challenges,” Shodh Sari-An International Multidisciplinary Journal, vol. 2, no. 4, pp. 328-343, 2023, doi: 10.59231/SARI7643.
F. T. H. Lo, “The paradoxical transparency of opaque machine learning,” AI & SOCIETY, vol. 39, no. 3, pp. 1397-1409, 2024, doi: 10.1007/s00146-022-01616-7.
E. SAHiN, N. N. Arslan, and D. Ozdemir, “Unlocking the black box: an in-depth review on interpreta-bility, explainability, and reliability in deep learning,” Neural Computing and Applications, vol. 37, no. 2, pp. 859-965, 2025, doi: 10.1007/s00521-024-10437-2.
O. Ilori, N. T. Nwosu, and H. N. N. Naiho, “Advanced data analytics in internal audits: A conceptual framework for comprehensive risk assessment and fraud detection,” Finance & Accounting Re-search Journal, vol. 6, no. 6, pp. 931-952, 2024, doi: 10.51594/farj.v6i6.1213.
D. Pelosi, D. Cacciagrano, and M. Piangerelli, “Explainability and Interpretability in Concept and Data Drift: A Systematic Literature Review,” Algorithms, vol. 18, no. 7, pp. 443-525, 2025, doi: 10.3390/a18070443.
D. Zhou and J. He, “Rare Category Analysis for Complex Data: A Review,” ACM Computing Surveys, vol. 56, no. 5, pp. 1-35, 2023, doi: 10.1145/3626520.
S. I. Kampezidou, A. Tikayat Ray, A. P. Bhat, O. J. Pinon Fischer, and D. N. Mavris, “Fundamental components and principles of su-pervised machine learning workflows with numerical and categorical data,” Eng, vol. 5, no. 1, pp. 384-416, 2024, doi: 10.3390/eng5010021.
Z. Wang, “Adaptive Ensemble Learning Framework with SHAP-Based Feature Optimization for Financial Anomaly Detection,” Artificial Intelligence and Machine Learning Review, vol. 5, no. 1, pp. 51-66, 2024, doi: 10.69987/AIMLR.2024.50105.
Y. Gebreyesus, D. Dalton, D. De Chiara, M. Chinnici, and A. Chinnici, “AI for automating data center operations: model ex-plainability in the data centre context using shapley additive explanations (SHAP),” Electronics, vol. 13, no. 9, pp. 1628-1647, 2024, doi: 10.3390/electronics13091628.
S. R. A. Parisineni and M. Pal, “Enhancing trust and interpretability of complex machine learning models using local interpretable model agnostic shap explanations,” International Journal of Data Science and Analytics, vol. 18, no. 4, pp. 457-466, 2024, doi: 10.1007/s41060-023-00458-w.
A. Chinnaraju, “Explainable AI (XAI) for trustworthy and transparent decision-making: A theoret-ical framework for AI interpretability,” World Journal of Advanced Engineering Technology and Sciences, vol. 14, no. 3, pp. 170-207, 2025, doi: 10.30574/wjaets.2025.14.3.0106.
T. A. Shaikh, T. Rasool, P. Verma, and W. A. Mir, “A fundamental overview of ensemble deep learning models and applications: systematic literature and state of the art,” Annals of Operations Research, no. 1, pp. 1-77, 2024, doi: 10.1007/s10479-024-06444-0.
M. Sakib, S. Mustajab, and M. Alam, “Ensemble deep learning techniques for time series analysis: a comprehensive review, applications, open issues, challenges, and future directions,” Cluster Computting, vol. 28, no. 1, pp. 73, 2025, doi: 10.1007/s10586-024-04684-0.
D. Zegarra Rodriguez, O. Daniel Okey, S. S. Maidin, E. Umoren Udo, and J. H. Kleinschmidt, “Attentive transformer deep learning al-gorithm for intrusion detection on IoT systems using automatic Xplainable feature selection,” Plos one, vol. 18, no. 10, Art. no. e0286652-e0286676, 2023, doi: 10.1371/journal.pone.0286652.
Abdullah, M. Ateeb Ather, O. Kolesnikova, and G. Sidorov, “Detection of Biased Phrases in the Wiki Neu-trality Corpus for Fairer Digital Content Management Using Artificial Intelligence,” Big Data and Cognitive Computing, vol. 9, no. 7, pp. 190-221, 2025, doi: 10.3390/bdcc9070190.
W. Liu, K. Liu, W. Sun, G. Yang, K. Ren, X. Meng, et al., “Self-supervised feature learning based on spectral masking for hyperspectral image classification,” IEEE Transactions on Geoscience and Remote Sensing, vol. 61, no. 1, pp. 1-15, 2023, doi: 10.1109/TGRS.2023.3310489.
Z. Huang, X. Jin, C. Lu, Q. Hou, M. M. Cheng, D. Fu, et al., “Contrastive masked autoencoders are stronger vision learners,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 46, no. 4, pp. 2506-2517, 2023, doi: 10.1109/TPAMI.2023.3336525.
N. Mazumdar and P. K. D. Sarma, “Sequential pattern mining algorithms and their applications: a technical review,” International Journal of Data Science and Analytics, vol. 20, no. 3, pp. 1683-1726, 2025, doi: 10.1007/s41060-024-00659-x.
S. Naz, M. H. Khan, and M. S. Farid, “Enhancing fall detection using multimodal time-series sensory data and VLAD encoding: a framework: S,” Naz et al. The Journal of Supercomputing, vol. 81, no. 8, pp. 890, 2025, doi: 10.1007/s11227-025-07370-z.
E. A. Asiamah, N. K. Akrasi-Mensah, P. Odame, E. Keelson, A. S. Agbemenu, E. T. Tchao, et al., “A storage-efficient learned indexing for blockchain systems using a sliding window search enhanced online gradient descent,” The Journal of Supercomputing, vol. 81, no. 1, pp. 1-29, 2025, doi: 10.1007/s11227-024-06805-3.
C. Chen, J. Liang, W. Sun, G. Yang, and X. Meng, “An automatically recursive feature elimination method based on threshold decision in random forest classification,” Geo-spatial information science, vol. 28, no. 4, pp. 1494-1519, 2025, doi: 10.1080/10095020.2024.2387457.
E. B. Gulcan and F. Can, “Unsupervised concept drift detection for multi-label data streams,” Artificial Intelligence Review, vol. 56, no. 3, pp. 2401-2434, 2023, doi: 10.1007/s10462-022-10232-2.
D. Lukats, O. Zielinski, A. Hahn, and F. Stahl, “A benchmark and survey of fully unsupervised concept drift detectors on real-world data streams,” International Journal of Data Science and Analytics, vol. 19, no. 1, pp. 1-31, 2025, doi: 10.1007/s41060-024-00620-y.
M. A. Belay, S. S. Blakseth, A. Rasheed, and P. Salvo Rossi, “Unsupervised anomaly detection for IoT-based multivariate time series: Existing solutions, performance analysis and future directions,” Sensors, vol. 23, no. 5, pp. 2844-2867, 2023, doi: 10.3390/s23052844.
Z. Zamanzadeh Darban, G. I. Webb, S. Pan, C. Aggarwal, and M. Salehi, “Deep learning for time series anomaly detection: A survey,” ACM Computing Surveys, vol. 57, no. 1, pp. 1-42, 2024, doi: 10.1145/3691338.
J. Yang and Q. Long, “A modification of adaptive moment estimation (adam) for machine learning,” Journal of Industrial and Management Optimization, vol. 20, no. 7, pp. 2516-2540, 2024, doi: 10.3934/jimo.2024014.
M. Reyad, A. M. Sarhan, and M. Arafa, “A modified Adam algorithm for deep neural network optimiza-tion,” Neural Computing and Applications, vol. 35, no. 23, pp. 17095-17112, 2023, doi: 10.1007/s00521-023-08568-z.