{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,10]],"date-time":"2026-07-10T02:23:57Z","timestamp":1783650237790,"version":"3.55.0"},"reference-count":69,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,10,16]],"date-time":"2025-10-16T00:00:00Z","timestamp":1760572800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>This study presents a machine learning (ML) approach for detecting non-compliance in companies' tax data. The dataset, consisting of over one million records, focuses on three key targets: invalid addresses, invalid director information, and invalid founder information. The analysis prioritizes young companies (\u2264 3 years old) with fewer than 100 employees, thereby improving class distributions and model effectiveness. A combination of binary classification techniques was employed, including benchmarked supervised learning models (XGBoost, Random Forest), anomaly detection methods (LOF, Isolation Forest), and semi-supervised learning using deep neural networks (DNNs) with unlabeled data. Given its computational efficiency, XGBoost was selected as the primary model. However, class imbalance persisted even among young companies, necessitating the integration of focal loss to improve classification performance. To further enhance accuracy while maintaining model interpretability, NSGA-II (Non-dominated Sorting Genetic Algorithm II) was used for multi-objective hyperparameter optimization of XGBoost. The objectives were to maximize ROC-AUC for improved predictive performance and minimize the number of trees to enhance interpretability. The optimized model achieved a ROC-AUC of 0.9417, compared to 0.9161 without optimization, demonstrating the effectiveness of this approach. Additionally, SHAP analysis provided insights into key factors influencing non-compliance, supporting explainability and aiding regulatory decision-making. This methodology contributes to fair and efficient oversight by reducing unnecessary inspections, minimizing disruptions to compliant businesses, and improving the overall effectiveness of tax compliance monitoring.<\/jats:p>","DOI":"10.3389\/frai.2025.1669191","type":"journal-article","created":{"date-parts":[[2025,10,16]],"date-time":"2025-10-16T05:31:21Z","timestamp":1760592681000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":4,"title":["Minimizing unnecessary tax audits using multi-objective hyperparameter tuning of XGBoost with focal loss"],"prefix":"10.3389","volume":"8","author":[{"given":"Ivan P.","family":"Malashin","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Igor S.","family":"Masich","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Vadim S.","family":"Tynchenko","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Andrei P.","family":"Gantimurov","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Vladimir A.","family":"Nelyub","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aleksei S.","family":"Borodulin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,10,16]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"100","DOI":"10.21070\/jas.v5i2.1530","article-title":"How significantly to emerging economies benefit from board attributes and risk management in enhancing firm profitability?","volume":"5","author":"Alabdullah","year":"2021","journal-title":"J. Account. Sci"},{"key":"B2","doi-asserted-by":"publisher","first-page":"18481","DOI":"10.1109\/ACCESS.2021.3052313","article-title":"Investigation of applying machine learning for watch-list filtering in anti-money laundering","volume":"9","author":"Alkhalili","year":"2021","journal-title":"IEEE Access"},{"key":"B3","doi-asserted-by":"publisher","first-page":"63","DOI":"10.1016\/j.bar.2005.09.001","article-title":"35 years of studies on business failure: an overview of the classic statistical methodologies and their related problems","volume":"38","author":"Balcaen","year":"2006","journal-title":"Br. Account. Rev"},{"key":"B4","doi-asserted-by":"publisher","first-page":"042","DOI":"10.30574\/wjarr.2024.23.1.1984","article-title":"Navigating financial compliance in small and medium-sized enterprises (smes): overcoming challenges and implementing effective solutions","volume":"23","author":"Bello","year":"2024","journal-title":"World J. Adv. Res. Rev"},{"key":"B5","first-page":"451","article-title":"\u201cArea under the precision-recall curve: point estimates and confidence intervals,\u201d","volume-title":"Joint European Conference on Machine Learning and Knowledge Discovery in Databases","author":"Boyd","year":"2013"},{"key":"B6","doi-asserted-by":"publisher","first-page":"2861","DOI":"10.1111\/1911-3846.12785","article-title":"Defining and managing corporate tax risk: perceptions of tax risk experts","volume":"39","author":"Br\u00fchne","year":"2022","journal-title":"Contemp. Account. Res"},{"key":"B7","doi-asserted-by":"publisher","first-page":"121464","DOI":"10.1016\/j.eswa.2023.121464","article-title":"Data-driven interpretable ensemble learning methods for the prediction of wind turbine power incorporating shap analysis","volume":"237","author":"Cakiroglu","year":"2024","journal-title":"Expert Syst. Appl"},{"key":"B8","doi-asserted-by":"publisher","first-page":"32","DOI":"10.1016\/j.neucom.2013.05.059","article-title":"A method for resampling imbalanced datasets in binary classification tasks for real-world problems","volume":"135","author":"Cateni","year":"2014","journal-title":"Neurocomputing"},{"key":"B9","doi-asserted-by":"publisher","first-page":"125216","DOI":"10.1016\/j.eswa.2024.125216","article-title":"Fiad: Graph anomaly detection framework based feature injection","volume":"259","author":"Chen","year":"2025","journal-title":"Expert Syst. Appl"},{"key":"B10","doi-asserted-by":"publisher","first-page":"182","DOI":"10.1109\/4235.996017","article-title":"A fast and elitist multiobjective genetic algorithm: Nsga-ii","volume":"6","author":"Deb","year":"2002","journal-title":"IEEE Trans. Evol. Comp"},{"key":"B11","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3643659.3643927","article-title":"\u201cAutomated boundary identification for machine learning classifiers,\u201d","volume-title":"Proceedings of the 17th ACM\/IEEE International Workshop on Search-Based and Fuzz Testing","author":"Dobslaw","year":"2024"},{"key":"B12","unstructured":"Tax penalties and tax compliance\n          \n          111\n          \n            \n              Doran\n              M.\n            \n          \n          Harv. J. Legis\n          46\n          2009"},{"key":"B13","doi-asserted-by":"publisher","first-page":"271","DOI":"10.1007\/s10506-015-9174-x","article-title":"Modelling compliance risk: a structured approach","volume":"23","author":"Esayas","year":"2015","journal-title":"Artif. Intellig. Law"},{"key":"B14","doi-asserted-by":"publisher","first-page":"657","DOI":"10.1111\/beer.12671","article-title":"The effect of government informatization construction on corporate digital technology innovation: new evidence from China","volume":"34","author":"Feng","year":"2024","journal-title":"Busin. Ethics Environ. Responsib"},{"key":"B15","unstructured":"\u201cCost-sensitive learning,\u201d\n          \n          \n            \n              Fern\u00e1ndez\n              A.\n            \n            \n              Garc\u00eda\n              S.\n            \n            \n              Galar\n              M.\n            \n            \n              Prati\n              R. C.\n            \n            \n              Krawczyk\n              B.\n            \n            \n              Herrera\n              F.\n            \n          \n          Learning From Imbalanced Data Sets\n          \n          2018"},{"key":"B16","doi-asserted-by":"publisher","first-page":"102405","DOI":"10.1016\/j.is.2024.102405","article-title":"Gama: A multi-graph-based anomaly detection framework for business processes via graph neural networks","volume":"124","author":"Guan","year":"2024","journal-title":"Inf. Syst"},{"key":"B17","doi-asserted-by":"crossref","first-page":"1674","DOI":"10.1109\/ICMLA52953.2021.00267","article-title":"\u201cAn empirical evaluation of the t-sne algorithm for data visualization in structural engineering,\u201d","volume-title":"2021 20th IEEE International Conference on Machine Learning and Applications (ICMLA)","author":"Hajibabaee","year":"2021"},{"key":"B18","doi-asserted-by":"publisher","first-page":"79","DOI":"10.1007\/s10115-017-1142-1","article-title":"Are we done with business process compliance: state of the art and challenges ahead","volume":"57","author":"Hashmi","year":"2018","journal-title":"Knowl. Inf. Syst"},{"key":"B19","doi-asserted-by":"publisher","first-page":"4806","DOI":"10.1109\/ACCESS.2019.2962617","article-title":"The real-world-weight cross-entropy loss function: Modeling the costs of mislabeling","volume":"8","author":"Ho","year":"2019","journal-title":"IEEE Access"},{"key":"B20","first-page":"5375","article-title":"\u201cLearning deep representation for imbalanced classification,\u201d","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition","author":"Huang","year":"2016"},{"key":"B21","doi-asserted-by":"publisher","first-page":"4360","DOI":"10.1016\/j.eswa.2014.01.012","article-title":"Topological pattern discovery and feature extraction for fraudulent financial reporting","volume":"41","author":"Huang","year":"2014","journal-title":"Expert Syst. Appl"},{"key":"B22","doi-asserted-by":"publisher","first-page":"763","DOI":"10.1016\/S0305-750X(99)00144-8","article-title":"A vat revenue simulation model for tax reform in developing countries","volume":"28","author":"Jenkins","year":"2000","journal-title":"World Dev"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.1098\/rsta.2009.0159","author":"Johnstone","year":"2009","journal-title":"Statistical challenges of high-dimensional"},{"key":"B24","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1080\/2573234X.2024.2438195","article-title":"A prediction model to detect non-compliant taxpayers using a supervised machine learning approach: evidence from tunisia","volume":"8","author":"Kamoun","year":"2024","journal-title":"J. Busin. Analyt"},{"key":"B25","doi-asserted-by":"publisher","first-page":"919","DOI":"10.1007\/s11292-022-09512-2","article-title":"False positives vs. false negatives: Public opinion on the cost ratio in criminal justice risk assessment","volume":"19","author":"Kang","year":"2023","journal-title":"J. Exp. Criminol"},{"key":"B26","doi-asserted-by":"publisher","first-page":"58","DOI":"10.1016\/j.knosys.2013.10.017","article-title":"A simple utility function with the rules-verified weights for analyzing the top competitiveness of wcy 2012","volume":"58","author":"Ko","year":"2014","journal-title":"Knowl.-Based Syst"},{"key":"B27","doi-asserted-by":"publisher","first-page":"100314","DOI":"10.1016\/j.bdr.2022.100314","article-title":"Downsampling for binary classification with a highly imbalanced dataset using active learning","volume":"28","author":"Lee","year":"2022","journal-title":"Big Data Res"},{"key":"B28","doi-asserted-by":"publisher","first-page":"7436","DOI":"10.3390\/s22197436","article-title":"An anomaly detection algorithm based on ensemble learning for 5G environment","volume":"22","author":"Lei","year":"2022","journal-title":"Sensors"},{"key":"B29","doi-asserted-by":"publisher","first-page":"80","DOI":"10.54254\/2755-2721\/2025.21736","article-title":"Anomaly detection in tax filing documents using natural language processing techniques","volume":"144","author":"Liang","year":"2025","journal-title":"Appl. Comp. Eng"},{"key":"B30","doi-asserted-by":"publisher","first-page":"122049","DOI":"10.1016\/j.eswa.2023.122049","article-title":"ASE: Anomaly scoring based ensemble learning for highly imbalanced datasets","volume":"238","author":"Liang","year":"2024","journal-title":"Expert Syst. Appl"},{"key":"B31","article-title":"\u201cHigh-level expert group on reforming the structure of the eu banking sector,\u201d","author":"Liikanen","year":"2012","journal-title":"Final Report, Brussels"},{"key":"B32","doi-asserted-by":"publisher","first-page":"849","DOI":"10.1109\/TVCG.2020.3030370","article-title":"Taxthemis: Interactive mining and exploration of suspicious tax evasion groups","volume":"27","author":"Lin","year":"2020","journal-title":"IEEE Trans. Vis. Comput. Graph"},{"key":"B33","doi-asserted-by":"publisher","first-page":"609","DOI":"10.1007\/s10479-021-04114-z","article-title":"Credit risk classification: an integrated predictive accuracy algorithm using artificial and deep neural networks","volume":"330","author":"Mahbobi","year":"2023","journal-title":"Ann. Operat. Res"},{"key":"B34","first-page":"840","article-title":"Predicting tax evasion using machine learning: a study of e-commerce transactions","volume":"3","author":"Mahesar","year":"2025","journal-title":"Spect. Eng. Sci"},{"key":"B35","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s00521-024-10323-x","article-title":"Automatically weighted focal loss for imbalance learning","volume":"2024","author":"Mahmoodi","year":"2024","journal-title":"Neural Comp. Appl"},{"key":"B36","doi-asserted-by":"publisher","first-page":"3011","DOI":"10.1007\/s40747-022-00641-9","article-title":"Deep learning model construction for a semi-supervised classification with feature learning","volume":"9","author":"Mandapati","year":"2023","journal-title":"Complex Intellig. Syst"},{"key":"B37","doi-asserted-by":"publisher","first-page":"233","DOI":"10.1108\/CG-09-2015-0122","article-title":"Boards attributes that increase firm risk-evidence from the uk","volume":"16","author":"Mathew","year":"2016","journal-title":"Corp. Govern"},{"key":"B38","doi-asserted-by":"publisher","first-page":"113128","DOI":"10.1016\/j.eswa.2019.113128","article-title":"Leveraging feature selection to detect potential tax fraudsters","volume":"145","author":"Matos","year":"2020","journal-title":"Expert Syst. Appl"},{"key":"B39","doi-asserted-by":"publisher","first-page":"176","DOI":"10.3390\/risks10090176","article-title":"The effect of corporate governance structure on fraud and money laundering","volume":"10","author":"Mousavi","year":"2022","journal-title":"Risks"},{"key":"B40","doi-asserted-by":"publisher","first-page":"731","DOI":"10.1080\/20421338.2023.2187930","article-title":"Predicting tax fraud using supervised machine learning approach","volume":"15","author":"Murorunkwere","year":"2023","journal-title":"Afr. J. Sci. Technol. Innovat. Dev"},{"key":"B41","doi-asserted-by":"publisher","first-page":"78658","DOI":"10.1109\/ACCESS.2021.3083060","article-title":"Machine learning for anomaly detection: a systematic review","volume":"9","author":"Nassif","year":"2021","journal-title":"IEEE Access"},{"key":"B42","doi-asserted-by":"publisher","first-page":"69896","DOI":"10.1109\/ACCESS.2022.3187837","article-title":"Dnn-based indoor localization under limited dataset using gans and semi-supervised learning","volume":"10","author":"Njima","year":"2022","journal-title":"IEEE Access"},{"key":"B43","unstructured":"Realistic evaluation of deep semi-supervised learning algorithms\n          \n          \n            \n              Oliver\n              A.\n            \n            \n              Odena\n              A.\n            \n            \n              Raffel\n              C. A.\n            \n            \n              Cubuk\n              E. D.\n            \n            \n              Goodfellow\n              I.\n            \n          \n          Adv. Neural Inf. Process. Syst\n          \n          2018"},{"key":"B44","doi-asserted-by":"publisher","first-page":"551","DOI":"10.1007\/s12599-020-00677-6","article-title":"Know-your-customer (kyc) requirements for initial coin offerings","volume":"63","author":"Ostern","year":"2021","journal-title":"Busin. Inform. Syst. Eng"},{"key":"B45","article-title":"\u201cCorporate risk taking and ownership structure,\u201d","author":"Paligorova","year":"2010","journal-title":"Technical Report, Bank of Canada Working Paper"},{"key":"B46","doi-asserted-by":"publisher","first-page":"457","DOI":"10.1007\/s41060-023-00458-w","article-title":"Enhancing trust and interpretability of complex machine learning models using local interpretable model agnostic shap explanations","volume":"18","author":"Parisineni","year":"2024","journal-title":"Int. J. Data Sci. Analyt"},{"key":"B47","doi-asserted-by":"crossref","first-page":"424","DOI":"10.1109\/SOLI.2013.6611452","article-title":"\u201cDoes a company has bright future? Predicting financial risk from revenue reports,\u201d","volume-title":"Proceedings of 2013 IEEE International Conference on Service Operations and Logistics, and Informatics","author":"Qian","year":"2013"},{"key":"B48","unstructured":"An exploration of artificial intelligence techniques for optimizing tax compliance, fraud detection, and revenue collection in modern tax administrations\n          \n          56\n          80\n          \n            \n              Rahman\n              S.\n            \n            \n              Sirazy\n              M.\n            \n            \n              Das\n              R.\n            \n            \n              Khan\n              R.\n            \n          \n          Int. J. Busin. Intellig. Big Data Analyt\n          7\n          2024"},{"key":"B49","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/ICAECT63952.2025.10958987","author":"Ramesh","year":"2025"},{"key":"B50","doi-asserted-by":"publisher","first-page":"102336","DOI":"10.1016\/j.sysarc.2021.102336","article-title":"Dnn pruning with principal component analysis and connection importance estimation","volume":"122","author":"Riera","year":"2022","journal-title":"J. Syst. Architect"},{"key":"B51","first-page":"2980","article-title":"\u201cFocal loss for dense object detection,\u201d","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition","author":"Ross","year":"2017"},{"key":"B52","doi-asserted-by":"publisher","first-page":"137","DOI":"10.1016\/S2212-5671(16)30025-9","article-title":"The effect of corporate governance components on return on assets and stock return of companies listed in tehran stock exchange","volume":"36","author":"Rostami","year":"2016","journal-title":"Procedia Econ. Finance"},{"key":"B53","doi-asserted-by":"publisher","first-page":"522","DOI":"10.1016\/j.ins.2021.05.055","article-title":"Approximating xgboost with an interpretable decision tree","volume":"572","author":"Sagi","year":"2021","journal-title":"Inf. Sci"},{"key":"B54","doi-asserted-by":"publisher","first-page":"728","DOI":"10.1108\/TQM-10-2020-0246","article-title":"The relationship between managerial attributes and firm risk-taking","volume":"34","author":"Salehi","year":"2022","journal-title":"TQM J"},{"key":"B55","doi-asserted-by":"publisher","first-page":"116409","DOI":"10.1016\/j.eswa.2021.116409","article-title":"Tax evasion risk management using a hybrid unsupervised outlier detection method","volume":"193","author":"Savi\u0107","year":"2022","journal-title":"Expert Syst. Appl"},{"key":"B56","doi-asserted-by":"publisher","first-page":"225","DOI":"10.1109\/TEVC.2014.2313407","article-title":"Visualization of pareto front approximations in evolutionary multiobjective optimization: a critical review and the prosection method","volume":"19","author":"Tu\u0161ar","year":"2014","journal-title":"IEEE Trans. Evolut. Comp"},{"key":"B57","doi-asserted-by":"publisher","first-page":"e0301541","DOI":"10.1371\/journal.pone.0301541","article-title":"Confirming the statistically significant superiority of tree-based machine learning algorithms over their counterparts for tabular data","volume":"19","author":"Uddin","year":"2024","journal-title":"PLoS ONE"},{"key":"B58","unstructured":"Third-party risks and liabilities in case of vat fraud in the eu\n          \n          31\n          \n            \n              Van Brederode\n              R. F.\n            \n          \n          Int'l Tax J\n          34\n          2008"},{"key":"B59","first-page":"11","article-title":"Visualizing data using t-SNE","volume":"9","author":"Van der Maaten","year":"2008","journal-title":"J. Mach. Learn. Res"},{"key":"B60","doi-asserted-by":"publisher","first-page":"57757","DOI":"10.1109\/ACCESS.2021.3070634","article-title":"A comprehensive review on nsga-ii for multi-objective combinatorial optimization problems","volume":"9","author":"Verma","year":"2021","journal-title":"IEEE Access"},{"key":"B61","doi-asserted-by":"publisher","first-page":"1534","DOI":"10.1109\/TPAMI.2007.70799","article-title":"Feature selection with kernel class separability","volume":"30","author":"Wang","year":"2008","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell"},{"key":"B62","doi-asserted-by":"publisher","first-page":"43","DOI":"10.1016\/j.aos.2015.11.005","article-title":"Client business models, process business risks and the risk of material misstatement of revenue","volume":"48","author":"Wright","year":"2016","journal-title":"Account. Organiz. Soc"},{"key":"B63","doi-asserted-by":"publisher","first-page":"107871","DOI":"10.1016\/j.ress.2021.107871","article-title":"Remaining useful life prediction based on intentional noise injection and feature reconstruction","volume":"215","author":"Xiao","year":"2021","journal-title":"Reliab. Eng. System Safety"},{"key":"B64","doi-asserted-by":"publisher","first-page":"137","DOI":"10.1007\/s10551-022-05120-2","article-title":"Using machine learning to predict corporate fraud: evidence based on the gone framework","volume":"186","author":"Xu","year":"2023","journal-title":"J. Busin. Ethics"},{"key":"B65","doi-asserted-by":"publisher","first-page":"100209","DOI":"10.1016\/j.jjimei.2023.100209","article-title":"AI-powered ensemble machine learning to optimize cost strategies in logistics business","volume":"4","author":"Yaiprasert","year":"2024","journal-title":"Int. J. Inform. Managem. Data Insights"},{"key":"B66","doi-asserted-by":"publisher","first-page":"105459","DOI":"10.1016\/j.frl.2024.105459","article-title":"Capital market liberalization and accounting quality: evidence from china","volume":"64","author":"Zhang","year":"2024","journal-title":"Finance Res. Letters"},{"key":"B67","doi-asserted-by":"publisher","first-page":"106898","DOI":"10.1016\/j.econmod.2024.106898","article-title":"Do sanctions imposed on peers have a deterrence effect on related-party transactions of observing firms? evidence from chinese listed firms","volume":"141","author":"Zhou","year":"","journal-title":"Econ. Model"},{"key":"B68","doi-asserted-by":"publisher","first-page":"11817","DOI":"10.1109\/TNNLS.2024.3389714","article-title":"Reconstructed graph neural network with knowledge distillation for lightweight anomaly detection","volume":"35","author":"Zhou","year":"","journal-title":"IEEE Trans. Neural Netw. Learn. Syst"},{"key":"B69","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3495161","article-title":"Graph neural networks: Taxonomy, advances, and trends","volume":"13","author":"Zhou","year":"2022","journal-title":"ACM Trans. Intellig. Syst. Technol"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1669191\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,16]],"date-time":"2025-10-16T05:31:28Z","timestamp":1760592688000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1669191\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,16]]},"references-count":69,"alternative-id":["10.3389\/frai.2025.1669191"],"URL":"https:\/\/doi.org\/10.3389\/frai.2025.1669191","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,10,16]]},"article-number":"1669191"}}