mlsanityData to 14 Oct 2026

Meta Two-Sample Testing: Learning Kernels for Testing with Limited Data

arXiv 2106.07636 ↗First posted 14 Jun 2021stat.ML cs.AI cs.LGcited by 18

Not on the map. Why?

Abstract (arXiv)

Modern kernel-based two-sample tests have shown great success in distinguishing complex, high-dimensional distributions with appropriate learned kernels. Previous work has demonstrated that this kernel learning procedure succeeds, assuming a considerable number of observed samples from each distribution. In realistic scenarios with very limited numbers of data samples, however, it can be challenging to identify a kernel powerful enough to distinguish complex distributions. We address this issue by introducing the problem of meta two-sample testing (M2ST), which aims to exploit (abundant) auxiliary data on related tasks to find an algorithm that can quickly identify a powerful test on new target tasks. We propose two specific algorithms for this task: a generic scheme which improves over baselines and a more tailored approach which performs even better. We provide both theoretical justification and empirical evidence that our proposed meta-testing schemes out-perform learning kernel-based tests directly from scarce observations, and identify when such schemes will be successful.

Key facts. Cited by 18 papers in the map (data through 14 Oct 2026). Builds on 32 of its 57 references that are in the map. First posted on arXiv 14 Jun 2021.

Builds on

32 of 57 references.

References · 32

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,897
  3. 2017Prototypical Networks for Few-shot Learningcited by 3,015
  4. 2015Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networkscited by 2,054
  5. 2019Do ImageNet Classifiers Generalize to ImageNet?cited by 1,515
  6. 2018Demystifying MMD GANscited by 1,293
  7. 2016Revisiting Classifier Two-Sample Testscited by 268
  8. 2018Learning to Propagate Labels: Transductive Propagation Network for Few-shot Learningcited by 211
  9. 2015On Wasserstein Two Sample Testing and Related Families of Nonparametric Testscited by 166
  10. 2020Learning Deep Kernels for Non-Parametric Two-Sample Testscited by 130
  11. 2016Generative Models and Model Criticism via Optimized Maximum Mean Discrepancycited by 100
  12. 2016Interpretable Distribution Features with Maximum Testing Powercited by 63
20 more
  1. 2015Fast Two-Sample Testing with Analytic Representations of Probability Measurescited by 53
  2. 2019Open Set Domain Adaptation: Theoretical Bound and Algorithmcited by 48
  3. 2016Classification accuracy as a proxy for two sample testingcited by 42
  4. 2017A Linear-Time Kernel Goodness-of-Fit Testcited by 41
  5. 2020Attribute Propagation Network for Graph Zero-shot Learningcited by 29
  6. 2018Robust Hypothesis Testing Using Wasserstein Uncertainty Setscited by 28
  7. 2019Two-sample Testing Using Deep Learningcited by 27
  8. 2013A new graph-based two-sample test for multivariate and object datacited by 24
  9. 2020CSCL: Critical Semantic-Consistent Learning for Unsupervised Domain Adaptationcited by 23
  10. 2020Learning Kernel Tests Without Data Splittingcited by 19
  11. 2021Isometric Propagation Network for Generalized Zero-shot Learningcited by 18
  12. 2020How does the Combined Risk Affect the Performance of Unsupervised Domain Adaptation Approaches?cited by 17
  13. 2021Neural Tangent Kernel Maximum Mean Discrepancycited by 15
  14. 2015Adaptivity and Computation-Statistics Tradeoffs for Kernel and Distance based High Dimensional Two Sample Testingcited by 14
  15. 2019Semantic-Transferable Weakly-Supervised Endoscopic Lesions Segmentationcited by 14
  16. 2021Sequential Domain Adaptation by Synthesizing Distributionally Robust Expertscited by 14
  17. 2019Noise Contrastive Meta-Learning for Conditional Density Estimation using Kernel Mean Embeddingscited by 10
  18. 2017Two-Sample Tests for Large Random Graphs Using Network Statisticscited by 7
  19. 2018Practical methods for graph two-sample testingcited by 7
  20. 2016Multivariate tests of association based on univariate testscited by 5

Built on by

Cited by year

Citing papers per year: 2022: 7; 2023: 3; 2024: 4; 2025: 4.

18 papers.

Sort:

  1. 2023MMD-FUSE: Learning and Combining Kernels for Two-Sample Testing Without Data Splittingcited by 33
  2. 2022Bilateral Dependency Optimization: Defending Against Model-inversion Attackscited by 30
  3. 2022Meta-learning Adaptive Deep Kernel Gaussian Processes for Molecular Property Predictioncited by 21
  4. 2022AutoML Two-Sample Testcited by 20
  5. 2024Detecting Machine-Generated Texts by Multi-Population Aware Optimization for Maximum Mean Discrepancycited by 20
  6. 2024A Unified Data Representation Learning for Non-parametric Two-sample Testingcited by 6
  7. 2025DUAL: Learning Diverse Kernels for Aggregated Two-sample and Independence Testingcited by 6
  8. 2023Diversity-enhancing Generative Network for Few-shot Hypothesis Adaptationcited by 4
  9. 2025Neural Network Reprogrammability: A Unified Theme on Model Reprogramming, Prompt Tuning, and Prompt Instructioncited by 4
  10. 2022AgraSSt: Approximate Graph Stein Statistics for Interpretable Assessment of Implicit Graph Generatorscited by 3
  11. 2025One Stone, Two Birds: Enhancing Adversarial Defense Through the Lens of Distributional Discrepancycited by 3
  12. 2022Adversarial Attack and Defense for Non-Parametric Two-Sample Testscited by 2
6 more
  1. 2023MLHOps: Machine Learning for Healthcare Operationscited by 2
  2. 2024Learning Representations for Independence Testingcited by 2
  3. 2022Multi-class Classification with Fuzzy-feature Observations: Theory and Algorithmscited by 1
  4. 2024Binary Hypothesis Testing for Softmax Models and Leverage Score Modelscited by 1
  5. 2025Anchor-based Maximum Discrepancy for Relative Similarity Testingcited by 1
  6. 2022RFFNet: Large-Scale Interpretable Kernel Methods via Random Fourier Featurescited by 0