└── README.md /README.md: -------------------------------------------------------------------------------- 1 | # Awesome Self-Supervised Learning for Non-Sequential Tabular Data (SSL4NSTD) 2 | ![Version](https://img.shields.io/badge/Version-1.0-lightgrey.svg) 3 | ![LastUpdated](https://img.shields.io/badge/LastUpdated-2025.01-lightblue.svg) 4 | ![Topic](https://img.shields.io/badge/Topic-SSL4NSTD-pink?logo=github) 5 | 6 | This repository contains the frontier research on **self-supervised learning** for tabular data which has been a popular topic recently.
7 | This list is maintained by [Wei-Wei Du](https://wwweiwei.github.io/) and [Wei-Yao Wang](https://github.com/wywyWang). (Actively keep updating)
8 | If you have come across relevant resources or found some errors in this repository, feel free to open an issue or submit a PR. 9 | 10 | ## Our Survey Paper 11 | [A Survey on Self-Supervised Learning for Non-Sequential Tabular Data (ACML-24 Journal Track)](https://arxiv.org/abs/2402.01204) 12 | ### Citation 13 | ```bibtex 14 | @article{DBLP:journals/corr/abs-2402-01204, 15 | author = {Wei{-}Yao Wang and 16 | Wei{-}Wei Du and 17 | Derek Xu and 18 | Wei Wang and 19 | Wen{-}Chih Peng}, 20 | title = {A Survey on Self-Supervised Learning for Non-Sequential Tabular Data}, 21 | journal = {CoRR}, 22 | volume = {abs/2402.01204}, 23 | year = {2024} 24 | } 25 | ``` 26 | 27 | ## Papers 28 | ### Predictive Learning 29 | * VIME: Extending the Success of Self- and Semi-supervised Learning to Tabular Domain (NeurIPS'20) [[Paper]](https://proceedings.neurips.cc/paper/2020/file/7d97667a3e056acab9aaf653807b4a03-Paper.pdf) [[Supplementary]](https://proceedings.neurips.cc/paper/2020/file/7d97667a3e056acab9aaf653807b4a03-Supplemental.pdf) [[Code]](https://github.com/jsyoon0823/VIME) 30 | * TaBERT: Pretraining for Joint Understanding of Textual and Tabular Data (ACL'20) [[Paper]](https://arxiv.org/abs/2005.08314) 31 | * TABBIE: Pretrained Representations of Tabular Data (NAACL'21) [[Paper]](https://arxiv.org/abs/2105.02584)) [[Code]](https://github.com/SFIG611/tabbie) 32 | * CORE: Self- and Semi-supervised Tabular Learning with COnditional REgularizations (NeurIPS'21) [[Paper]](https://sslneurips21.github.io/files/CameraReady/CORE_workshop.pdf) 33 | * TabTransformer: Tabular Data Modeling Using Contextual Embeddings [[Paper]](https://arxiv.org/abs/2012.06678) 34 | * TabNet: Attentive Interpretable Tabular Learning (AAAI'21) [[Paper]](https://arxiv.org/abs/1908.07442) [Code](https://github.com/dreamquark-ai/tabnet) 35 | * Self-Supervision Enhanced Feature Selection with Correlated Gates (ICLR'22) [[Paper]](https://openreview.net/pdf?id=oDFvtxzPOx) [[Code]](https://github.com/chl8856/SEFS) 36 | * TransTab: Learning Transferable Tabular Transformers Across Tables (NeurIPS'22) [[Paper]](https://arxiv.org/abs/2205.09328) [[Code]](https://github.com/RyanWangZf/transtab) [[Blog]](https://realsunlab.medium.com/transtab-learning-transferable-tabular-transformers-across-tables-1e34eec161b8) 37 | * LIFT: Language-Interfaced Fine-Tuning for Non-language Machine Learning Tasks (NeurIPS'22) [[Paper]](https://arxiv.org/pdf/2206.06565.pdf) [[Code]](https://github.com/UW-Madison-Lee-Lab/LanguageInterfacedFineTuning) 38 | * Self Supervised Pre-training for Large Scale Tabular Data (NeurIPS'22 TRL Workshop) [[Paper]](https://table-representation-learning.github.io/assets/papers/self_supervised_pre_training_f.pdf) [[Blog]](https://www.amazon.science/publications/self-supervised-pre-training-for-large-scale-tabular-data) 39 | * Local Contrastive Feature Learning for Tabular Data (CIKM'22) [[Paper]](https://dl.acm.org/doi/pdf/10.1145/3511808.3557630) 40 | * Revisiting Self-Training with Regularized Pseudo-Labeling for Tabular Data (preprint'23) [[Paper]](https://arxiv.org/abs/2302.14013) 41 | * Generative Table Pre-training Empowers Models for Tabular Prediction (EMNLP'23) [[Paper]](https://arxiv.org/pdf/2305.09696.pdf) [[Code]](https://github.com/ZhangTP1996/TapTap) 42 | * TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second (ICLR'23) [[Paper]](https://arxiv.org/pdf/2207.01848.pdf) [[Code]](https://github.com/automl/TabPFN) 43 | * STUNT: Few-shot Tabular Learning with Self-generated Tasks from Unlabeled Tables (ICLR'23) [[Paper]](https://arxiv.org/pdf/2303.00918.pdf) [[Code]](https://github.com/jaehyun513/STUNT) 44 | * Language Models are Realistic Tabular Data Generators (ICLR'23) [[Paper]](https://arxiv.org/pdf/2210.06280.pdf) [[Code]](https://github.com/kathrinse/be_great) 45 | * Self-supervised Representation Learning from Random Data Projectors (NeurIPS'23 TRL Workshop) [[Paper]](https://arxiv.org/pdf/2310.07756.pdf) [[Code]](https://github.com/layer6ai-labs/lfr) 46 | * SwitchTab: Switched Autoencoders Are Effective Tabular Learners (AAAI'24) [[Paper]](https://arxiv.org/pdf/2401.02013.pdf) 47 | * Making Pre-trained Language Models Great on Tabular Prediction (ICLR'24) [[Paper]](https://openreview.net/pdf?id=anzIzGZuLi) 48 | * Binning as a Pretext Task: Improving Self-Supervised Learning in Tabular Domains (ICML'24) [[Paper]](https://arxiv.org/abs/2405.07414) [[Code]](https://github.com/kyungeun-lee/tabularbinning) 49 | * Large Scale Transfer Learning for Tabular Data via Language Modeling (NeurIPS'24) [[Paper]](https://arxiv.org/abs/2406.12031) [[Code]](https://github.com/mlfoundations/rtfm) 50 | * Accurate predictions on small data with a tabular foundation model (Nature-25) [[Paper]](https://www.nature.com/articles/s41586-024-08328-6) 51 | 52 | ### Contrastive Learning 53 | * SCARF: Self-Supervised Contrastive Learning using Random Feature Corruption (ICLR'22) [[Paper]](https://arxiv.org/pdf/2106.15147.pdf) [[Code]](https://github.com/clabrugere/pytorch-scarf) 54 | * STab: Self-supervised Learning for Tabular Data (NeurIPS'22 Workshop on TRL) [[Paper]](https://openreview.net/pdf?id=EfR55bFcrcI) 55 | * TransTab: Learning Transferable Tabular Transformers Across Tables (NeurIPS'22) [[Paper]](https://arxiv.org/pdf/2205.09328.pdf) 56 | * PTaRL: Prototype-based Tabular Representation Learning via Space Calibration (ICLR'24) [[Paper]](https://openreview.net/pdf?id=G32oY4Vnm8) 57 | 58 | ### Hybrid Learning 59 | * SubTab: Subsetting Features of Tabular Data for Self-Supervised Representation Learning (NeurIPS'21) [[Paper]](https://arxiv.org/pdf/2110.04361.pdf) [[Supplementary]](https://openreview.net/attachment?id=vrhNQ7aYSdr&name=supplementary_material) [[Code]](https://github.com/AstraZeneca/SubTab) 60 | * SAINT: Improved Neural Networks for Tabular Data via Row Attention and Contrastive Pre-Training (NurIPS‘22 Workshop on TRL) [[Paper]](https://arxiv.org/pdf/2106.01342.pdf) [[Code]](https://github.com/somepago/saint) 61 | * Transfer Learning with Deep Tabular Models (ICLR'23) [[Paper]](https://arxiv.org/pdf/2206.15306.pdf) [[Code]](https://github.com/LevinRoman/tabular-transfer-learning) 62 | * DoRA: Domain-Based Self-Supervised Learning Framework for Low-Resource Real Estate Appraisal (CIKM'23) [[Paper]](https://arxiv.org/abs/2309.00855) [[Code]](https://github.com/wwweiwei/DoRA) 63 | * ReConTab: Regularized Contrastive Representation Learning for Tabular Data (NeurIPS'23 Workshop on TRL) [[Paper]](https://arxiv.org/pdf/2310.18541.pdf) 64 | * XTab: Cross-table Pretraining for Tabular Transformers (ICML'23) [[Paper]](https://arxiv.org/abs/2305.06090) 65 | * UniTabE: A Universal Pretraining Protocol for Tabular Foundation Model in Data Science (ICLR'24) [[Paper]](https://arxiv.org/pdf/2307.09249.pdf) 66 | 67 | ## Benchmarks 68 | | Benchmark | Task | #Datasets | Paper | 69 | |--------------|---------------------------------------|-----------|-------| 70 | | MLPCBench | Classification | 40 | [Kadra et al., 2021](https://arxiv.org/abs/2106.11189) | 71 | | DLBench | Classification, Regression | 11 | [Shwartz-Ziv and Armon, 2022](https://arxiv.org/abs/2106.03253) | 72 | | TabularBench | Classification, Regression | 45 | [Grinsztajn et al., 2022](https://arxiv.org/abs/2207.08815) | 73 | | TabZilla | Classification | 36 | [McElfresh et al., 2023](https://arxiv.org/abs/2305.02997) | 74 | | TabPretNet | Unlabeled, Classification, Regression | 2000 | [Ye et al., 2023](https://arxiv.org/abs/2307.04308) | 75 | | The Tremendous TabLib Trawl (T4) | Unlabeled | 3.1M | [Gardner et al., 2024](https://arxiv.org/abs/2406.12031v1) | 76 | 77 | ## Tutorials 78 | * Self-Supervised Learning: Self-Prediction and Contrastive Learning (NeurIPS'21) [[Website]](https://neurips.cc/virtual/2021/tutorial/21895) 79 | 80 | ## Workshops 81 | * Table Representation Learning (NeurIPS) [[Website]](https://table-representation-learning.github.io/) 82 | 83 | ## Related Survey 84 | * Deep Neural Networks and Tabular Data: A Survey [[Paper]](https://arxiv.org/abs/2110.01889) 85 | * Self-Supervised Learning for Recommender Systems: A Survey (TKDE) [[Paper]](https://arxiv.org/pdf/2203.15876.pdf) 86 | * Beyond Just Vision: A Review on Self-Supervised Representation Learning on Multimodal and Temporal Data [[Paper]](https://arxiv.org/abs/2206.02353) 87 | * Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects [[Paper]](https://arxiv.org/abs/2306.10125) 88 | * On the Opportunities and Challenges of Foundation Models for Geospatial Artificial Intelligence [[Paper]](https://arxiv.org/abs/2304.06798) 89 | * A Survey on Time-Series Pre-Trained Models [[Paper]](https://arxiv.org/abs/2305.10716) 90 | 91 | ## Tools & Libraries 92 | * Pytorch Frame: A modular deep learning framework for building neural network models on heterogeneous tabular data [[Link]](https://github.com/pyg-team/pytorch-frame#implemented-deep-tabular-models) 93 | * PyTorch Tabular: A Framework for Deep Learning with Tabular Data [[Link]](https://github.com/manujosephv/pytorch_tabular) 94 | * Pytorch wide-deep: A flexible package for multimodal-deep-learning to combine tabular data with text and images using Wide and Deep models in Pytorch [[Link]](https://github.com/jrzaurin/pytorch-widedeep) 95 | --------------------------------------------------------------------------------