A Review of Tabular Data Understanding and Automated Analytics

Author

Priyanka Parate, Kapur Pardhi, Sanskruti Raut , Ayan Sheikh ,Mrs. Umme Ayeman Saqib Gani

Abstract

 Reliable analysis of tabular data requires decisions about column meaning, data quality, model suitability and the interpretation of results. Progress in predictive modeling alone does not resolve these dependencies. This narrative review synthesizes 36 selected works spanning data profiling, semantic type detection, automated machine learning, tree ensembles, neural predictors, tabular foundation models, natural language interfaces, visualization and explainability. The comparison shows that tree ensembles remain useful baselines, while improved neural defaults, efficient ensembling and in-context learning broaden the available accuracy–cost trade-offs. Recent work on feature engineering and uncertainty also shows why predictive rankings cannot stand in for complete workflow evaluation. Existing AutoML systems already automate substantial parts of preparation and model construction; the unresolved challenge is to connect these capabilities to explicit semantic evidence, validated user intent and auditable explanations. The review identifies evaluation requirements for such integration without claiming a universally superior model or experimentally validating a new system.

Keywords

tabular data; semantic profiling; automated machine learning; foundation models; conversational analytics; explainable artificial intelligence.

Full Text:

Download Paper PDF

References

[1] H. Chen, R. H. L. Chiang, and V. C. Storey, “Business intelligence and analytics: From big data to big impact,” MIS Quarterly, vol. 36, no. 4, pp. 1165–1188, 2012.

[2] Z. Abedjan, L. Golab, and F. Naumann, “Profiling relational data: A survey,” The VLDB Journal, vol. 24, pp. 557–581, 2015.

[3] F. Hutter, L. Kotthoff, and J. Vanschoren, Eds., Automated Machine Learning: Methods, Systems, Challenges. Springer, 2019.

[4] M. Hulsebos et al., “Sherlock: A deep learning approach to semantic data type detection,” in Proc. KDD, pp. 1500–1508, 2019.

[5] D. Zhang, Y. Suhara, J. Li, M. Hulsebos, Ç. Demiralp, and W.-C. Tan, “Sato: Contextual semantic type detection in tables,” Proc. VLDB Endowment, vol. 13, no. 12, pp. 1835–1848, 2020.

[6] A. Tschalzev et al., “TabPrep: Closing the feature engineering gap in tabular benchmarks,” arXiv:2606.02384, 2026, preprint version.

[7] L. Breiman, “Random forests,” Machine Learning, vol. 45, pp. 5–32, 2001.

[8] T. Chen and C. Guestrin, “XGBoost: A scalable tree boosting system,” in Proc. KDD, pp. 785–794, 2016.

[9] G. Ke et al., “LightGBM: A highly efficient gradient boosting decision tree,” in Advances in Neural Information Processing Systems, vol. 30, 2017.

[10] L. Prokhorenkova et al., “CatBoost: Unbiased boosting with categorical features,” in Advances in Neural Information Processing Systems, vol. 31, 2018.

[11] L. Grinsztajn, E. Oyallon, and G. Varoquaux, “Why do tree-based models still outperform deep learning on typical tabular data?” in Advances in Neural Information Processing Systems, vol. 35, 2022.

[12] Y. Gorishniy, I. Rubachev, V. Khrulkov, and A. Babenko, “Revisiting deep learning models for tabular data,” in Advances in Neural Information Processing Systems, vol. 34, 2021.

[13] D. Holzmüller, L. Grinsztajn, and I. Steinwart, “Better by default: Strong pre-tuned MLPs and boosted trees on tabular data,” in Advances in Neural Information Processing Systems, vol. 37, 2024.

[14] Y. Gorishniy, A. Kotelnikov, and A. Babenko, “TabM: Advancing tabular deep learning with parameter-efficient ensembling,” in Proc. ICLR, 2025.

[15] B. Zhu, X. Shi, N. Erickson, M. Li, G. Karypis, and M. Shoaran, “XTab: Cross-table pretraining for tabular transformers,” in Proc. ICML, PMLR, vol. 202, pp. 43181–43204, 2023.

[16] N. Hollmann et al., “Accurate predictions on small data with a tabular foundation model,” Nature, vol. 637, pp. 319–326, 2025.

[17] J. Qu, D. Holzmüller, G. Varoquaux, and M. Le Morvan, “TabICL: A tabular foundation model for in-context learning on large data,” in Proc. ICML, PMLR, vol. 267, pp. 50817–50847, 2025.

[18] J. Qu, D. Holzmüller, G. Varoquaux, and M. Le Morvan, “TabICLv2: A better, faster, scalable, and open tabular foundation model,” in Proc. ICML, PMLR, vol. 306, pp. 102477–102555, 2026.

[19] M. Spinaci, M. Polewczyk, M. Schambach, and S. Thelin, “ConTextTab: A semantics-aware tabular in-context learner,” in Advances in Neural Information Processing Systems, vol. 38, 2025.

[20] J. Heredge, M. J. Villani, P. Deshpande, A. Seshadri, and N. Kumar, “CRUMB: Efficient prior fitted network inference via distributionally matched context batching,” arXiv:2606.11473, 2026, preprint.

[21] J. Robertson, A. Reuter, S. Guo, N. Hollmann, F. Hutter, and B. Schölkopf, “Do-PFN: In-context learning for causal effect estimation,” in Advances in Neural Information Processing Systems, vol. 38, 2025.

[22] B. Pfahringer, H. Bensusan, and C. Giraud-Carrier, “Meta-learning by landmarking various learning algorithms,” in Proc. ICML, pp. 743–750, 2000.

[23] J. Vanschoren, “Meta-learning,” in Automated Machine Learning, F. Hutter, L. Kotthoff, and J. Vanschoren, Eds., pp. 35–61, 2019.

[24] C. Thornton, F. Hutter, H. H. Hoos, and K. Leyton-Brown, “Auto-WEKA: Combined selection and hyperparameter optimization of classification algorithms,” in Proc. KDD, pp. 847–855, 2013.

[25] M. Feurer et al., “Efficient and robust automated machine learning,” in Advances in Neural Information Processing Systems, vol. 28, pp. 2962–2970, 2015.

[26] N. Erickson et al., “AutoGluon-Tabular: Robust and accurate AutoML for structured data,” arXiv:2003.06505, 2020.

[27] C. Yu, Y. Jin, Q. Xing, Y. Zhang, S. Guo, and S. Meng, “Advanced user credit risk prediction model using LightGBM, XGBoost and Tabnet with SMOTEENN,” arXiv:2408.03497, 2024.

[28] N. Erickson et al., “TabArena: A living benchmark for machine learning on tabular data,” in Advances in Neural Information Processing Systems, vol. 38, 2025.

[29] A. Vaswani et al., “Attention is all you need,” in Advances in Neural Information Processing Systems, vol. 30, pp. 5998–6008, 2017.

[30] T. B. Brown et al., “Language models are few-shot learners,” in Advances in Neural Information Processing Systems, vol. 33, pp. 1877–1901, 2020.

[31] T. Yu et al., “Spider: A large-scale human-labeled dataset for complex and cross-domain semantic parsing and text-to-SQL task,” in Proc. EMNLP, pp. 3911–3921, 2018.

[32] P. Lewis et al., “Retrieval-augmented generation for knowledge-intensive NLP tasks,” in Advances in Neural Information Processing Systems, vol. 33, pp. 9459–9474, 2020.

[33] K. Wongsuphasawat et al., “Voyager: Exploratory analysis via faceted browsing of visualization recommendations,” IEEE Transactions on Visualization and Computer Graphics, vol. 22, no. 1, pp. 649–658, 2016.

[34] M. T. Ribeiro, S. Singh, and C. Guestrin, “Why should I trust you? Explaining the predictions of any classifier,” in Proc. KDD, pp. 1135–1144, 2016.

[35] S. M. Lundberg and S.-I. Lee, “A unified approach to interpreting model predictions,” in Advances in Neural Information Processing Systems, vol. 30, 2017.

[36] J. L. De Melo Costa, F. Popineau, A. Rimmel, and B.-L. Doan, “High performance, low reliability: Uncertainty benchmarking for tabular foundation models,” in Proc. ESANN, pp. 115–120, 2026.

Share your valuable work from Social Media Buttons