Pith. sign in

REVIEW 4 major objections 4 minor 300 references

SiamJEPA: On the Role of Siamese Student Encoders in JEPA

T0 review · 4 major / 4 minor · reviewed 2026-08-02 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Giving a JEPA model two student encoders and aligning the two views improves its representations and speeds early learning.

desk verdict Useful ablation inside a two-branch model, but the headline claim about single-encoder JEPA is not supported by the experiments. read the letter →

arxiv 2607.04044 v2 pith:555AWMEC submitted 2026-07-04 cs.CV stat.ML

Makoto Yamada This is my paper · ORCID
classification cs.CVstat.ML
keywords self-supervisedlearningJEPASiamesestudentencodersmaskedlatentpredictionEMAteacherKLregularizationImageNetlinearprobingvisiontransformer
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper asks what two student encoders contribute to a joint-embedding predictive architecture (JEPA), and answers that they act as a regularizer for the latent-prediction objective. SiamJEPA feeds two disjointly masked views of an image to a weight-shared student pair, predicts the EMA teacher's masked-token latents, and adds a KL term that aligns probabilistic latents inferred from the two views. The paper reports that raising this KL term improves ImageNet linear probing from 68.91% to 70.15% at 400 epochs, and accelerates early training: at epoch 101 the regularized model reaches 63.44% versus 60.88%. If right, the result matters because Siamese consistency becomes a tunable inductive bias that other JEPA-style models could adopt without changing the prediction task.

What carries the argument

The load-bearing object is the Siamese student pair with its KL-regularized probabilistic predictor. Two masked views produce two latent representations; a posterior distribution is built from both views, a prior from only one, and the KL loss (with stop-gradient on the prior) makes each single view predictive of the information shared across views. λKL continuously interpolates between the Siamese configuration and a JEPA-like configuration, and the EMA teacher provides the latent targets for the normalized-MSE masked-prediction term (Sim-2). Disjoint masking prevents shortcut solutions by ensuring the two views are complementary.

What would settle it

Train a genuine single-encoder JEPA with the same codebase, data, batch size, and 400-epoch budget, and evaluate it with the same linear probe. If it reaches or outperforms SiamJEPA's 70.15% without the second student encoder, then the Siamese encoders are not the source of the gain; conversely, reproducing 70.15% only when the KL term is active would support the paper's regularizer claim.

Watch

Extended reading notes

Core claim

SiamJEPA is a masked latent-prediction architecture in which two student encoders with shared weights process independently and disjointly masked views of the same image. An EMA teacher produces targets for masked-token prediction, while a KL term (Sim-1) forces the probabilistic latent variable inferred from both views to match the prior inferred from each single view, with a stop-gradient on the prior. The central claim is that this KL consistency term is a regularizer, not a nuisance: with λKL=0.00001 the model behaves like a JEPA-style baseline, and raising λKL to 0.01 improves ImageNet linear probing from 68.91% to 70.15% after 400 epochs and accelerates early-stage learning (63.44% vs

Load-bearing premise

The central comparison assumes SiamJEPA with λKL=0.00001 behaves as a JEPA-like baseline, yet the two Siamese branches and the KL term remain in the model; no true single-encoder JEPA is trained, so the paper's claim that Siamese encoders beat single-encoder JEPA depends on an equivalence asserted in §3.3 and Table 2 rather than demonstrated.

Editorial extensions

If this is right

  • If the KL term is a genuine regularizer, turning it up should keep improving representation separability up to a point; the paper finds λKL values of 0.01–0.03 beat 0.00001 across all training lengths.
  • The early-training gap (63.44% vs 60.88% at epoch 101) means a Siamese-regularized JEPA can reach a given linear-probe accuracy in roughly half the epochs of its weakly regularized counterpart.
  • Larger weight decay becomes useful for SiamJEPA under long training, while the weakly regularized baseline plateaus or collapses, suggesting the Siamese regularizer interacts positively with other forms of regularization.
  • Because the KL term is decoupled from the masked-prediction loss, the recipe should transfer to other JEPA-style architectures and modalities if the paper's inductive-bias claim is correct.
  • Even simple random masking becomes competitive with block masking when the mask ratio is tuned, indicating that the Siamese consistency, not the masking sophistication, drives the gain.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper never trains a true single-encoder JEPA in its own codebase; a direct run of such a model would settle whether the reported gain comes from having two encoders or from the KL regularizer working within a two-encoder setup.
  • If the KL term is mainly a regularizer, simpler alignment objectives, such as cosine similarity between pooled view representations, might reproduce much of the gain at lower computational cost; this is a natural testable extension.
  • The largest gap appears early in training, so the benefit may be most valuable at short pretraining budgets or in continual-learning settings where rapid representation formation matters.
  • The paper itself flags that its findings may not transfer to larger backbones or deeper predictors; testing SiamJEPA at larger scale and with deeper predictors is the open check implied by its own discussion.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes SiamJEPA, a JEPA-style self-supervised model with two Siamese student encoders processing disjoint masked views, an EMA teacher encoder, a KL consistency term between the two student branches, and a latent prediction objective with a probabilistic predictor. The method is evaluated by ImageNet-1K linear probing with ViT-Base, using the MAE codebase as a common framework. The central empirical claim is that Siamese student encoders act as a regularizer: increasing the KL weight from λKL=1e-5 to 0.01 improves linear probing accuracy and accelerates early learning, and the model reaches about 70.15% at 400 epochs versus 68.91% for the JEPA-like baseline. The paper also compares against MAE, CAE, and I-JEPA, and reports ablations of mask type, weight decay, free-bit threshold, learning rate, pooling strategy, EMA schedule, and predictor depth.

Significance. If the claims are supported, the paper would provide a useful empirical study of an architectural inductive bias in latent predictive models, with the attractive practical finding that a small KL-consistency term improves representation quality and convergence speed within a two-branch architecture. The work is reproducible in spirit: the code URL is given, the implementation is based on the public MAE codebase, and the ablations are extensive, including per-epoch linear probing curves. However, the headline architectural claim — that Siamese student encoders outperform ``single-encoder JEPA variants'' — is not tested, because no single-encoder JEPA is trained. The internal comparison of λKL values is consistent and supports a more modest claim about KL regularization strength, but the paper's abstract and conclusion assert a broader result.

major comments (4)
  1. [§3.3, Table 2] The central comparison to ``single-encoder JEPA'' is not made. The baseline labeled JEPA-like is SiamJEPA with λKL=0.00001, which still contains two Siamese student encoders and, crucially, the posterior q(Z(1)|h(H_cls(1)), H_cls(2)) defined in §3.2 conditions on both branches and feeds both into the predictor g. Even as λKL→0, the model does not reduce to a single-encoder JEPA; the posterior coupling remains. The sentence in §3.3 that λKL=0 makes the model ``similar to that of single encoder JEPA'' is therefore asserted, not demonstrated. Since no I-JEPA-style single-encoder model is trained in the same codebase, the abstract's ``consistently outperforms comparable single-encoder JEPA variants'' and the conclusion's ``corresponding JEPA-like baseline'' overstate the evidence. The paper should either train a genuine single-encoder baseline (e.g., removing the second branch and conditioni
  2. [Table 2 vs Tables 3 and 7, §4.1] The configuration of the headline SiamJEPA result is internally inconsistent. §4.1 states that all SiamJEPA models are evaluated with mean pooling of the 10th layer, but Table 2's note says the JEPA-like baseline uses mean pooling of the 10th layer while SiamJEPA uses mean pooling of the final (12th) layer. More importantly, Table 2 reports 70.7% for SiamJEPA with λKL=0.01 and weight decay 0.05 at epoch 400, but Table 3 (λKL=0.010, weight decay 0.05, 10th layer) reports 69.33%, and Table 7 (weight decay 0.1) reports 70.15% for the 10th layer and 70.71% for the 12th layer. The value 70.7 does not match any clearly specified row. The authors must harmonize the evaluation protocol and report the exact configuration (layer, pooling, weight decay) for every number in Table 2.
  3. [§4.2, Table 1] The training-efficiency comparison with MAE conflates epochs with compute. Table 1 shows MAE uses an effective batch size of 4096 while SiamJEPA uses 8192, so at 400 epochs SiamJEPA processes twice as many images. The paper's statement that SiamJEPA outperforms MAE ``using less than one-quarter of the training epochs'' ignores this difference. If the claim is about limited training budgets, the comparison should report wall-clock time or FLOPs, or match the total number of samples processed. As written, the efficiency claim is not fully supported.
  4. [Table 3, §4.3] The strongest comparison — 70.15% for λKL=0.01 versus 68.91% for λKL=0.00001 — also changes weight decay: the λKL=0.01 run uses weight decay 0.1 while the weak-KL baseline uses weight decay 0.05. Table 3 does include a λKL=0.00001, weight-decay-0.1 run, but it collapses to 62.54%, so the interaction is real. However, the paper should make this interaction explicit in the text and in Table 2, since the headline gap is not purely attributable to λKL. A clean same-weight-decay comparison at 0.05 gives 69.33% versus 68.91%, a much smaller difference, which should be reported in the main comparison so readers can see the effect size.
minor comments (4)
  1. [§4.5] The text refers to ``Figure 5 shows the ablation study of masking strategies,'' but the actual results are in Table 5. The reference should be corrected.
  2. [§3.3] The text says ``SiamJEPA model with λKL = 10−4 as a JEPA-like method,'' but the tables use λKL = 10−5 (0.00001). This is a factor-of-ten discrepancy and should be fixed.
  3. [§3.1] There is a duplicated ``where where'' in the decomposition of the encoder outputs. Minor typo.
  4. [§4.9] The predictor-depth discussion is honestly hedged, but the conclusion in §5 still says the Siamese architecture ``serves as an effective regularizer.'' The open questions in §4.9 and §6 (small scale, sensitivity to implementation details, predictor-depth uncertainty) should be reflected more prominently in the abstract, which currently states the architectural claim without qualification.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central empirical claims are benchmark measurements and do not reduce by construction to the paper's equations or to its self-citations.

full rationale

The paper's central claim—that Siamese student encoders act as a regularizer and improve JEPA representation quality—is supported by ImageNet linear probing results that are internally measured, not derived from the loss equations. The comparison is an ablation over lambda_KL in a fixed two-branch SiamJEPA architecture: the final loss L = 1/2(MSE^(1)+MSE^(2)) + lambda_KL/2(KL_sg^(1)+KL_sg^(2)) is not shown to reduce to another objective by substitution, and the reported accuracy numbers are empirical runs rather than fitted constants renamed as predictions. The main concern in the manuscript is not circularity but construct validity: the assertion that lambda_KL=0 makes the model 'similar to that of single encoder JEPA' is unsupported, since the two Siamese branches and the posterior coupling q(Z^(1)|h(H_cls^(1)), H_cls^(2)) remain even when lambda_KL is near zero. That is a factual/claim-support problem, not a case where the result is forced by definition or by a self-citation chain. The repeated citations to PhiNet and Yamada et al. are architectural provenance and hyperparameter heuristics, not external uniqueness theorems or machine-checked results that carry the burden of the empirical conclusion. The ImageNet evaluation is an external benchmark, so the central claim remains independent of the prior work cited. No circular step meeting the required evidentiary standard is present.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

No new physical entities are postulated. The latent variable Z and probabilistic predictors are standard modeling components borrowed from prior work; the only new objects are hyperparameters and architectural variants of existing Siamese/EMA systems.

free parameters (7)
  • KL regularization weight λKL = 0.01 / 0.03 (ablated 0.00001)
    Strength of the Siamese consistency loss; chosen after linear-probe ablations in Table 3. Central to the regularizer claim.
  • Free-bit threshold = 0.1 default, 0.05 best in ablation
    Caps the KL term to avoid collapse; selected by ablation in Table 4.
  • EMA momentum schedule = 0.99 → 0.999 → 0.9999
    Schedule is empirically determined and changed at epochs 200 and 300; Section 4.8 notes it was not optimized.
  • Weight decay = 0.1 (ablated 0.05)
    Larger weight decay helps longer training; Table 3 and Section 4.4.
  • Mask ratio = 0.75 block masking
    Selected among 0.7, 0.75, 0.8 based on Table 5; best performance.
  • Base learning rate = 1.5e-4
    Best of 1.0e-4, 1.5e-4, 2.0e-4, 2.5e-4 in Table 6.
  • Decoder depth = 1
    Shallow predictor empirically better, Section 4.9.
assumptions (5)
  • domain assumption EMA teacher prevents representation collapse
    Borrowed from BYOL/DINO/I-JEPA without proof; assumed throughout.
  • domain assumption Disjoint mask sets M1∩M2=∅ prevent shortcut learning
    Stated as crucial in Section 3.1, not validated by an ablation comparing overlap vs disjoint masks.
  • domain assumption KL regularization with free-bit prevents collapse while allowing useful discrepancy
    Invoked in Section 4.3 to justify the regularization design.
  • ad hoc to paper Stop-gradient on the prior branch improves stability and performance
    Introduced as a preliminary-experiment heuristic in Section 3.3, attributed to Yamada et al. 2025.
  • domain assumption The masked prediction objective (Sim-2) supplies the primary learning signal
    Inferred from loss curves in Section 4.3, not proven; central to the interpretation that the KL term is a regularizer.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SiamJEPA: On the Role of Siamese Student Encoders in JEPA." pith.science (2026). https://pith.science/paper/555AWMEC

@misc{pith2026260704044,
  author       = {Pith},
  title        = {Pith review of: SiamJEPA: On the Role of Siamese Student Encoders in JEPA},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/555AWMEC}},
  note         = {Machine review of arXiv:2607.04044}
}
read the original abstract

Recently, Joint Embedding Predictive Architectures (JEPAs) have attracted significant attention in the computer vision and machine learning communities as a promising framework for self-supervised representation learning. Unlike masked autoencoders that reconstruct pixels, JEPA models learn representations by predicting latent embeddings of masked regions. Existing JEPA-based methods, such as I-JEPA and V-JEPA, typically employ a single encoder in the student network. In contrast, using Siamese encoders for student network is more naturally aligned with brain-inspired representation learning frameworks, yet their role in JEPA models remains largely unexplored. In this paper, we investigate the effect of Siamese student encoders in JEPA-based representation learning. To this end, we propose SiamJEPA, masked Siamese student encoders equipped with an exponential moving average (EMA) teacher network. SiamJEPA can also be viewed as a JEPA formulation of the brain-inspired representation learning model PhiNet. Through extensive experiments on ImageNet linear probing, we demonstrate that Siamese encoders act as an effective regularizer for the JEPA objective, improving representation separability and accelerating learning during the early stages of training. Furthermore, SiamJEPA consistently outperforms comparable single-encoder JEPA variants under limited training budgets and achieves higher linear probing accuracy than Masked Autoencoders (MAE) which requires longer training. Our findings reveal that Siamese student encoders are not merely an architectural choice but constitute an important inductive bias for predictive representation learning. These results provide new insights into the design of JEPA-based models and suggest that incorporating Siamese student architectures offers a simple yet effective approach for improving self-supervised representation learning.

Figures

Figures reproduced from arXiv: 2607.04044 by the authors.

Figure 1
Figure 1. JEPA and SiamJEPA architectures. Sim-1 is a loss function to align the Siamese encoders. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Learning curves for λKL = 0.00001 and λKL = 0.01. With a small regularization coefficient, the KL divergence between the representations produced by the two Siamese student encoders remains large. In contrast, with a larger regularization coefficient, the KL divergence quickly converges to the free-bit threshold (0.1 in our experiments). The linear probing performance follows a similar trend. The training loss start… view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

300 extracted references · 24 linked inside Pith

  1. [1]

    Foundations of computational mathematics , volume =

    Gromov--Wasserstein distances and the metric approach to object matching , author =. Foundations of computational mathematics , volume =. 2011 , publisher =

    2011

  2. [2]

    arXiv preprint arXiv:2506.09985 , year=

    V-jepa 2: Self-supervised video models enable understanding, prediction and planning , author=. arXiv preprint arXiv:2506.09985 , year=

    arXiv

  3. [3]

    arXiv preprint arXiv:2404.08471 , year=

    Revisiting feature prediction for learning visual representations from video , author=. arXiv preprint arXiv:2404.08471 , year=

    arXiv

  4. [4]

    CVPR , year=

    Self-supervised learning from images with a joint-embedding predictive architecture , author=. CVPR , year=

  5. [5]

    arXiv preprint arXiv:2508.10104 , year=

    Dinov3 , author=. arXiv preprint arXiv:2508.10104 , year=

    arXiv

  6. [6]

    arXiv:1807.03748 , year =

    Representation learning with contrastive predictive coding , author =. arXiv:1807.03748 , year =

    arXiv

  7. [7]

    ICML , year =

    Fair and Optimal Classification via Post-Processing , author =. ICML , year =

  8. [8]

    Transactions on Machine Learning Research , year =

    Costs and Benefits of Fair Regression , author =. Transactions on Machine Learning Research , year =

Show all 300 references
  1. [9]

    CVPR , year =

    Understanding and constructing latent modality structures in multi-modal representation learning , author =. CVPR , year =

  2. [10]

    Computationally Efficient

    Toyokuni, Ayato and Yokoi, Sho and Kashima, Hisashi and Yamada, Makoto , booktitle =. Computationally Efficient

  3. [11]

    CVPR , year =

    Arcface: Additive angular margin loss for deep face recognition , author =. CVPR , year =

  4. [12]

    AIChE journal , volume =

    Nonlinear principal component analysis using autoassociative neural networks , author =. AIChE journal , volume =. 1991 , publisher =

    1991

  5. [13]

    arXiv preprint arXiv:1312.6114 , year =

    Auto-encoding variational bayes , author =. arXiv preprint arXiv:1312.6114 , year =

    arXiv

  6. [14]

    Nature , volume =

    Self-organizing neural network that discovers surfaces in random-dot stereograms , author =. Nature , volume =. 1992 , publisher =

    1992

  7. [15]

    arXiv preprint arXiv:2204.00616 , year =

    Simplicial embeddings in self-supervised learning and downstream classification , author =. arXiv preprint arXiv:2204.00616 , year =

    arXiv

  8. [16]

    ICML , year =

    A simple framework for contrastive learning of visual representations , author =. ICML , year =

  9. [17]

    TMLR , year =

    Approximating 1-Wasserstein Distance with Trees , author =. TMLR , year =

  10. [18]

    Foundations and Trends

    Concentration of measure inequalities in information theory, communications, and coding , author =. Foundations and Trends. 2013 , publisher =

    2013

  11. [19]

    IEEE Transactions on Big Data , volume =

    Billion-scale similarity search with gpus , author =. IEEE Transactions on Big Data , volume =. 2019 , publisher =

    2019

  12. [20]

    NeurIPS , year =

    Sublinear time low-rank approximation of distance matrices , author =. NeurIPS , year =

  13. [21]

    Journal of the ACM (JACM) , volume =

    Low-rank approximation and regression in input sparsity time , author =. Journal of the ACM (JACM) , volume =. 2017 , publisher =

    2017

  14. [22]

    ICLR , year =

    Hyperbolic neural networks++ , author =. ICLR , year =

  15. [23]

    IEEE Transactions on Neural Networks , volume =

    Fast and robust fixed-point algorithms for independent component analysis , author =. IEEE Transactions on Neural Networks , volume =. 1999 , publisher =

    1999

  16. [24]

    Advances in neural information processing systems , volume =

    Hyperbolic neural networks , author =. Advances in neural information processing systems , volume =

  17. [25]

    ICML , year =

    Hyperbolic entailment cones for learning hierarchical embeddings , author =. ICML , year =

  18. [26]

    International conference on machine learning , pages =

    Learning continuous hierarchies in the lorentz model of hyperbolic geometry , author =. International conference on machine learning , pages =. 2018 , organization =

    2018

  19. [27]

    The handbook of brain theory and neural networks , volume =

    Convolutional networks for images, speech, and time series , author =. The handbook of brain theory and neural networks , volume =. 1995 , publisher =

    1995

  20. [28]

    Advances in Neural Information Processing Systems , volume =

    Tree! i am no tree! i am a low dimensional hyperbolic embedding , author =. Advances in Neural Information Processing Systems , volume =

  21. [29]

    ICML , year =

    Lorentzian distance learning for hyperbolic representations , author =. ICML , year =

  22. [30]

    SIAM review , volume =

    Finding structure with randomness: Probabilistic algorithms for constructing approximate matrix decompositions , author =. SIAM review , volume =. 2011 , publisher =

    2011

  23. [31]

    arXiv preprint arXiv:1810.06546 , year =

    Poincar 'e glove: Hyperbolic word embeddings , author =. arXiv preprint arXiv:1810.06546 , year =

    arXiv

  24. [32]

    COLT , year =

    Sample-optimal low-rank approximation of distance matrices , author =. COLT , year =

  25. [33]

    arXiv preprint arXiv:1409.0473 , year =

    Neural machine translation by jointly learning to align and translate , author =. arXiv preprint arXiv:1409.0473 , year =

    arXiv

  26. [34]

    NIPS , year =

    Attention is all you need , author =. NIPS , year =

  27. [35]

    Nature Neuroscience , volume =

    Hippocampal spatial representations exhibit a hyperbolic geometry that expands with experience , author =. Nature Neuroscience , volume =. 2023 , publisher =

    2023

  28. [36]

    ICLR , year =

    Graph-based Nearest Neighbor Search in Hyperbolic Spaces , author =. ICLR , year =

  29. [37]

    arXiv preprint arXiv:2009.00836 , year =

    Nearest neighbor search for hyperbolic embeddings , author =. arXiv preprint arXiv:2009.00836 , year =

    2009 arXiv

  30. [38]

    Nickel, Maximillian and Kiela, Douwe , journal =. Poincar

  31. [39]

    1977 , publisher =

    The numerical treatment of integral equations , author =. 1977 , publisher =

    1977

  32. [40]

    ICML , year =

    Supervised tree-wasserstein distance , author =. ICML , year =

  33. [41]

    AISTATS , year =

    Wasserstein regularization for sparse multi-task regression , author =. AISTATS , year =

  34. [42]

    AISTATS , year =

    Fixed Support Tree-Sliced Wasserstein Barycenter , author =. AISTATS , year =

  35. [43]

    NeurIPS , year =

    Tree-sliced approximation of wasserstein distances , author =. NeurIPS , year =

  36. [44]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =

    The phylogenetic Kantorovich--Rubinstein metric for environmental sequence samples , author =. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =. 2012 , publisher =

    2012

  37. [45]

    Exploring simple

    Chen, Xinlei and He, Kaiming , booktitle=. Exploring simple

  38. [46]

    ICCV , year=

    Emerging properties in self-supervised vision transformers , author=. ICCV , year=

  39. [47]

    NeurIPS , year=

    Unsupervised learning of visual features by contrasting cluster assignments , author=. NeurIPS , year=

  40. [48]

    CVPR , year =

    Sliced Wasserstein kernels for probability distributions , author =. CVPR , year =

  41. [49]

    International journal of computer vision , volume =

    The earth mover's distance as a metric for image retrieval , author =. International journal of computer vision , volume =. 2000 , publisher =

    2000

  42. [50]

    ICLR , year =

    Sliced Wasserstein auto-encoders , author =. ICLR , year =

  43. [51]

    ICML , year =

    Optimal transport kernels for sequential and parallel neural architecture search , author =. ICML , year =

  44. [52]

    NIPS , year =

    Stochastic optimization for large-scale optimal transport , author =. NIPS , year =

  45. [53]

    A method for solving the convex programming problem with convergence rate O (1/k\^

    Nesterov, Yurii E , booktitle =. A method for solving the convex programming problem with convergence rate O (1/k\^

  46. [54]

    NIPS , year =

    Principal differences analysis: Interpretable characterization of differences between distributions , author =. NIPS , year =

  47. [55]

    Journal of the American Statistical Association , volume =

    Optimal estimation of Wasserstein distance on a tree with an application to microbiome studies , author =. Journal of the American Statistical Association , volume =. 2021 , publisher =

    2021

  48. [56]

    SIAM journal on imaging sciences , volume =

    A fast iterative shrinkage-thresholding algorithm for linear inverse problems , author =. SIAM journal on imaging sciences , volume =. 2009 , publisher =

    2009

  49. [57]

    Theoretical computer science , volume =

    Clustering to minimize the maximum intercluster distance , author =. Theoretical computer science , volume =. 1985 , publisher =

    1985

  50. [58]

    Applied and environmental microbiology , volume =

    UniFrac: a new phylogenetic method for comparing microbial communities , author =. Applied and environmental microbiology , volume =. 2005 , publisher =

    2005

  51. [59]

    NeurIPS , year =

    Online sinkhorn: Optimal transport distances from sample streams , author =. NeurIPS , year =

  52. [60]

    AISTATS , year =

    Proximal Optimal Transport Modeling of Population Dynamics , author =. AISTATS , year =

  53. [61]

    ICML , year =

    Improving mini-batch optimal transport via partial transportation , author =. ICML , year =

  54. [62]

    Foundations and Trends

    Computational optimal transport , author =. Foundations and Trends. 2019 , publisher =

    2019

  55. [63]

    Proceedings of the American Mathematical Society , volume =

    Diagonal Equivalence to Matrices with Prescribed Row and Column Sums , author =. Proceedings of the American Mathematical Society , volume =

  56. [64]

    IEEE Transactions on Pattern Analysis and Machine Intelligence , volume =

    Cross-domain matching with squared-loss mutual information , author =. IEEE Transactions on Pattern Analysis and Machine Intelligence , volume =. 2015 , publisher =

    2015

  57. [65]

    JMLR , volume =

    Multiple kernel learning algorithms , author =. JMLR , volume =

  58. [66]

    NIPS , year =

    Sinkhorn distances: Lightspeed computation of optimal transport , author =. NIPS , year =

  59. [67]

    AISTATS , year =

    Cross-domain object matching with model selection , author =. AISTATS , year =

  60. [68]

    Journal of the Royal Statistical Society

    A general class of coefficients of divergence of one distribution from another , author =. Journal of the Royal Statistical Society. Series B (Methodological) , pages =. 1966 , publisher =

    1966

  61. [69]

    2012 , publisher =

    Elements of information theory , author =. 2012 , publisher =

    2012

  62. [70]

    Proceedings of the Fourth Berkeley Symposium on Mathematical Statistics and Probability, Volume 1: Contributions to the Theory of Statistics , year =

    On measures of entropy and information , author =. Proceedings of the Fourth Berkeley Symposium on Mathematical Statistics and Probability, Volume 1: Contributions to the Theory of Statistics , year =

  63. [71]

    AISTATS , year =

    On the estimation of alpha-divergences , author =. AISTATS , year =

  64. [72]

    Advances in Applied Probability , volume =

    Integral probability metrics and their generating classes of functions , author =. Advances in Applied Probability , volume =. 1997 , publisher =

    1997

  65. [73]

    2008 , publisher =

    Optimal transport: old and new , author =. 2008 , publisher =

    2008

  66. [74]

    2000 , publisher =

    Probability for statisticians , author =. 2000 , publisher =

    2000

  67. [75]

    ISIT , year =

    Jensen-Shannon divergence and Hilbert space embedding , author =. ISIT , year =

  68. [76]

    2012 , publisher =

    Nonparametric and semiparametric models , author =. 2012 , publisher =

    2012

  69. [77]

    Neural computation , volume =

    Relative density-ratio estimation for robust distribution comparison , author =. Neural computation , volume =. 2013 , publisher =

    2013

  70. [78]

    ICML , year =

    A Swiss Army Knife for Minimax Optimal Transport , author =. ICML , year =

  71. [79]

    arXiv preprint arXiv:2006.07260 , year =

    Handling Multiple Costs in Optimal Transport: Strong Duality and Efficient Computation , author =. arXiv preprint arXiv:2006.07260 , year =

    2006 arXiv

  72. [80]

    arXiv preprint arXiv:1704.00805 , year =

    On the properties of the softmax function with application in game theory and reinforcement learning , author =. arXiv preprint arXiv:1704.00805 , year =

    arXiv

  73. [81]

    CVPR , year =

    Semantic Correspondence as an OPtimal Transport Problem , author =. CVPR , year =

  74. [82]

    AISTATS , year =

    Structured optimal transport , author =. AISTATS , year =

  75. [83]

    NeurIPS , year =

    Hierarchical optimal transport for document representation , author =. NeurIPS , year =

  76. [84]

    NeurIPS , year =

    Fast Unbalanced Optimal Transport on Tree , author =. NeurIPS , year =

  77. [85]

    AISTATS , year =

    Unsupervised Hierarchy Matching with Optimal Transport over Hyperbolic Spaces , author =. AISTATS , year =

  78. [86]

    NIPS , year =

    Kernel Feature Selection via Conditional Covariance Minimization , author =. NIPS , year =

  79. [87]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =

    Model selection and estimation in regression with grouped variables , author =. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =. 2006 , publisher =

    2006

  80. [88]

    , author =

    Semi-Supervised Optimal Transport for Heterogeneous Domain Adaptation. , author =. IJCAI , year =

  81. [89]

    CVPR , year =

    Semantic correspondence as an optimal transport problem , author =. CVPR , year =

  82. [90]

    arXiv preprint arXiv:1909.02373 , year =

    LSMI-Sinkhorn: Semi-supervised Squared-Loss Mutual Information Estimation with Optimal Transport , author =. arXiv preprint arXiv:1909.02373 , year =

    1909 arXiv

  83. [91]

    ICML , year =

    Subspace Robust Wasserstein Distances , author =. ICML , year =

  84. [92]

    Naval research logistics quarterly , volume =

    An algorithm for quadratic programming , author =. Naval research logistics quarterly , volume =. 1956 , publisher =

    1956

  85. [93]

    , author =

    Revisiting Frank-Wolfe: Projection-free sparse convex optimization. , author =. ICML , year =

  86. [94]

    Journal of Machine Learning Research , volume =

    A kernel two-sample test , author =. Journal of Machine Learning Research , volume =

  87. [95]

    ICLR , year =

    Post selection inference with incomplete maximum mean discrepancy estimator , author =. ICLR , year =

  88. [96]

    Advances in neural information processing systems , pages =

    Generative adversarial nets , author =. Advances in neural information processing systems , pages =

  89. [97]

    NIPS , year =

    B-test: A non-parametric, low variance kernel two-sample test , author =. NIPS , year =

  90. [98]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =

    Sure independence screening for ultrahigh dimensional feature space , author =. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =. 2008 , publisher =

    2008

  91. [99]

    arXiv preprint arXiv:1602.07358 , year =

    Post-selection inference for L1-penalized likelihood models , author =. arXiv preprint arXiv:1602.07358 , year =

    arXiv

  92. [100]

    The Annals of Statistics , volume =

    Exact post-selection inference, with application to the lasso , author =. The Annals of Statistics , volume =. 2016 , publisher =

    2016

  93. [101]

    arXiv preprint arXiv:1506.07997 , year =

    An Efficient Post-Selection Inference on High-Order Interaction Models , author =. arXiv preprint arXiv:1506.07997 , year =

    arXiv

  94. [102]

    Annals of statistics , volume =

    A significance test for the lasso , author =. Annals of statistics , volume =. 2014 , publisher =

    2014

  95. [103]

    NeurIPS , year =

    Generalized sliced wasserstein distances , author =. NeurIPS , year =

  96. [104]

    AISTATS , year =

    Smooth and Sparse Optimal Transport , author =. AISTATS , year =

  97. [105]

    Mathematical programming , volume =

    Smooth minimization of non-smooth functions , author =. Mathematical programming , volume =. 2005 , publisher =

    2005

  98. [106]

    ICML , year =

    Regularized Optimal Transport is Ground Cost Adversarial , author =. ICML , year =

  99. [107]

    Max-Sliced

    Deshpande, Ishan and Hu, Yuan-Ting and Sun, Ruoyu and Pyrros, Ayis and Siddiqui, Nasir and Koyejo, Sanmi and Zhao, Zhizhen and Forsyth, David and Schwing, Alexander G , booktitle =. Max-Sliced

  100. [108]

    NIPS , year =

    A kernel statistical test of independence , author =. NIPS , year =

  101. [109]

    The annals of applied statistics , volume =

    Random lasso , author =. The annals of applied statistics , volume =

  102. [110]

    Statistics and Computing , pages =

    Large-Scale Kernel Methods for Independence Testing , author =. Statistics and Computing , pages =

  103. [111]

    JMLR , volume =

    Large scale multiple kernel learning , author =. JMLR , volume =

  104. [112]

    Multiple kernel learning, conic duality, and the

    Bach, Francis R and Lanckriet, Gert RG and Jordan, Michael I , booktitle =. Multiple kernel learning, conic duality, and the

  105. [113]

    JMLR , volume =

    Generalized Hierarchical Kernel Learning , author =. JMLR , volume =

  106. [114]

    NIPS , year =

    Learning non-linear combinations of kernels , author =. NIPS , year =

  107. [115]

    IJCNN , year =

    Hand movement recognition for brazilian sign language: a study using distance-based neural networks , author =. IJCNN , year =

  108. [116]

    NIPS , year =

    Exploring large feature spaces with hierarchical multiple kernel learning , author =. NIPS , year =

  109. [117]

    Journal of classification , volume =

    Comparing partitions , author =. Journal of classification , volume =. 1985 , publisher =

    1985

  110. [118]

    The Annals of Statistics , volume =

    Component selection and smoothing in multivariate nonparametric regression , author =. The Annals of Statistics , volume =. 2006 , publisher =

    2006

  111. [119]

    NeurIPS , year =

    Bootstrap your own latent-a new approach to self-supervised learning , author =. NeurIPS , year =

  112. [120]

    arXiv preprint arXiv:2003.04297 , year =

    Improved baselines with momentum contrastive learning , author =. arXiv preprint arXiv:2003.04297 , year =

    2003 arXiv

  113. [121]

    arXiv preprint arXiv:2505.11129 , year=

    Brain-Inspired Stochastic Joint Embedding Representation Learning , author=. arXiv preprint arXiv:2505.11129 , year=

    arXiv

  114. [122]

    ICML , year =

    Zbontar, Jure and Jing, Li and Misra, Ishan and LeCun, Yann and Deny, St. ICML , year =

  115. [123]

    arXiv preprint arXiv:2104.13712 , year =

    A note on connecting barlow twins with negative-sample-free contrastive learning , author =. arXiv preprint arXiv:2104.13712 , year =

    arXiv

  116. [124]

    CVPR , year =

    Momentum contrast for unsupervised visual representation learning , author =. CVPR , year =

  117. [125]

    NIPS , year =

    Learning with a Wasserstein loss , author =. NIPS , year =

  118. [126]

    ICML , year =

    Multiple indefinite kernel learning with mixed norm regularization , author =. ICML , year =

  119. [127]

    Technical University of Denmark , volume =

    The matrix cookbook , author =. Technical University of Denmark , volume =

  120. [128]

    NIPS , year =

    Imagenet classification with deep convolutional neural networks , author =. NIPS , year =

  121. [129]

    IEEE Signal Processing Magazine , volume =

    Deep neural networks for acoustic modeling in speech recognition: The shared views of four research groups , author =. IEEE Signal Processing Magazine , volume =. 2012 , publisher =

    2012

  122. [130]

    Bioinformatics , volume =

    Genome-wide association analysis by lasso penalized logistic regression , author =. Bioinformatics , volume =. 2009 , publisher =

    2009

  123. [131]

    Shoemaker, Robert H , journal =. The

  124. [132]

    2006 , journal =

    Lamb, Justin and others , title =. 2006 , journal =

    2006

  125. [133]

    ECML , year =

    Bayesian multi-view tensor factorization , author =. ECML , year =

  126. [134]

    Efficient and robust feature selection via joint _

    Nie, Feiping and Huang, Heng and Cai, Xiao and Ding, Chris H , booktitle =. Efficient and robust feature selection via joint _

  127. [135]

    Exclusive Feature Learning on Arbitrary Structures via _

    Kong, Deguang and Fujimaki, Ryohei and Liu, Ji and Nie, Feiping and Ding, Chris , booktitle =. Exclusive Feature Learning on Arbitrary Structures via _

  128. [136]

    An Introduction to the Art of Mathematical Inequalities: The

    Steele, J Michael , year =. An Introduction to the Art of Mathematical Inequalities: The

  129. [137]

    IJCAI , year =

    Change-point detection with feature selection in high-dimensional time-series data , author =. IJCAI , year =

  130. [138]

    KDD , year =

    Gradient boosted feature selection , author =. KDD , year =

  131. [139]

    PASCAL Workshop on Statistics and Optimization of Clustering Workshop , year =

    Convex clustering shrinkage , author =. PASCAL Workshop on Statistics and Optimization of Clustering Workshop , year =

  132. [140]

    ACMMM , year =

    Towards multi-semantic image annotation with graph regularized exclusive group lasso , author =. ACMMM , year =

  133. [141]

    arXiv preprint arXiv:1601.04586 , year =

    Sparse Convex Clustering , author =. arXiv preprint arXiv:1601.04586 , year =

    arXiv

  134. [142]

    arXiv preprint arXiv:1304.0499 , year =

    Splitting methods for convex clustering , author =. arXiv preprint arXiv:1304.0499 , year =

    arXiv

  135. [143]

    ICML , year =

    Clusterpath an algorithm for clustering using convex fusion penalties , author =. ICML , year =

  136. [144]

    Applied and Computational Harmonic Analysis , volume =

    Sparse regression using mixed norms , author =. Applied and Computational Harmonic Analysis , volume =. 2009 , publisher =

    2009

  137. [145]

    , author =

    Efficient Generalized Fused Lasso and its Application to the Diagnosis of Alzheimer's Disease. , author =. AAAI , year =

  138. [146]

    arXiv preprint arXiv:1502.02009 , year =

    A General Analysis of the Convergence of ADMM , author =. arXiv preprint arXiv:1502.02009 , year =

    arXiv

  139. [147]

    , author =

    Feature Selection with Linked Data in Social Media. , author =. SDM , year =

  140. [148]

    arXiv preprint arXiv:1509.06397 , year =

    SnapVX: A Network-Based Convex Optimization Solver , author =. arXiv preprint arXiv:1509.06397 , year =

    arXiv

  141. [149]

    , author =

    Exclusive Lasso for Multi-task Feature Selection. , author =. AISTATS , year =

  142. [150]

    ECML , year =

    Higher Order Fused Regularization for Supervised Learning with Grouped Parameters , author =. ECML , year =

  143. [151]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =

    Sparsity and smoothness via the fused lasso , author =. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =. 2005 , publisher =

    2005

  144. [152]

    KDD , year =

    Network lasso: Clustering and optimization in large graphs , author =. KDD , year =

  145. [153]

    AISTATS , year =

    A Scalable Algorithm for Structured Kernel Feature Selection , author =. AISTATS , year =

  146. [154]

    Journal of Computational Biology , volume =

    Graphlet kernels for prediction of functional residues in protein structures , author =. Journal of Computational Biology , volume =. 2010 , publisher =

    2010

  147. [155]

    ISCIS , year =

    Fast Correlation Based Filter (FCBF) with a different search strategy , author =. ISCIS , year =

  148. [156]

    ICML , year =

    Feature selection for high-dimensional data: A fast correlation-based filter solution , author =. ICML , year =

  149. [157]

    SDM , year =

    Transductive HSIC Lasso , author =. SDM , year =

  150. [158]

    JMLR , volume =

    Conditional likelihood maximisation: a unifying framework for information theoretic feature selection , author =. JMLR , volume =. 2012 , publisher =

    2012

  151. [159]

    JMLR , volume =

    Towards Ultrahigh Dimensional Feature Selection for Big Data , author =. JMLR , volume =

  152. [160]

    JMLR , volume =

    Local causal and markov blanket induction for causal discovery and feature selection for classification part i: Algorithms and empirical evaluation , author =. JMLR , volume =

  153. [161]

    2011 , publisher =

    Spectral feature selection for data mining , author =. 2011 , publisher =

    2011

  154. [162]

    Annals of Statistics , pages =

    Greedy function approximation: a gradient boosting machine , author =. Annals of Statistics , pages =. 2001 , publisher =

    2001

  155. [163]

    PLoS computational biology , volume =

    Predicting positive p53 cancer rescue regions using Most Informative Positive (MIP) active learning , author =. PLoS computational biology , volume =. 2009 , publisher =

    2009

  156. [164]

    and Sriperumbudur, B

    Balasubramanian, K. and Sriperumbudur, B. and Lebanon, G. , booktitle =. Ultrahigh Dimensional Feature Screening via

  157. [165]

    Approximate

    Morup, M and Madsen, Kristoffer Hougaard and Hansen, Lars Kai , booktitle =. Approximate

  158. [166]

    Neural computation , volume =

    High-Dimensional Feature Selection by Feature-Wise Kernelized Lasso , author =. Neural computation , volume =. 2014 , publisher =

    2014

  159. [167]

    The Annals of statistics , volume =

    Least angle regression , author =. The Annals of statistics , volume =. 2004 , publisher =

    2004

  160. [168]

    , author =

    Efficient Spectral Feature Selection with Minimum Redundancy. , author =. AAAI , year =

  161. [169]

    Signal Processing Letters, IEEE , volume =

    Dual-augmented Lagrangian method for efficient sparse reconstruction , author =. Signal Processing Letters, IEEE , volume =. 2009 , publisher =

    2009

  162. [170]

    JMLR , volume =

    Feature selection via dependence maximization , author =. JMLR , volume =. 2012 , publisher =

    2012

  163. [171]

    Biometrika , volume =

    Notes on the history of correlation , author =. Biometrika , volume =. 1920 , publisher =

    1920

  164. [172]

    Annals of statistics , volume =

    Variable selection in nonparametric additive models , author =. Annals of statistics , volume =. 2010 , publisher =

    2010

  165. [173]

    Proceedings of the National Academy of Sciences , volume =

    Regulation of gene expression in the mammalian eye and its relevance to eye disease , author =. Proceedings of the National Academy of Sciences , volume =. 2006 , publisher =

    2006

  166. [174]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =

    Stability selection , author =. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =. 2010 , publisher =

    2010

  167. [175]

    Bolasso:

    Bach, Francis , booktitle =. Bolasso:. 2008 , organization =

    2008

  168. [176]

    The Annals of Statistics , year =

    Fast learning rate of multiple kernel learning: Trade-off between sparsity and smoothness , author =. The Annals of Statistics , year =

  169. [177]

    Signal Processing , volume =

    Semi-supervised speaker identification under covariate shift , author =. Signal Processing , volume =. 2010 , publisher =

    2010

  170. [178]

    JMLR , volume =

    Efficient online and batch learning using forward backward splitting , author =. JMLR , volume =. 2009 , publisher =

    2009

  171. [179]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =

    The group lasso for logistic regression , author =. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =. 2008 , publisher =

    2008

  172. [180]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =

    Regularization and variable selection via the elastic net , author =. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =. 2005 , publisher =

    2005

  173. [181]

    and Mohri, M

    Cortes, C. and Mohri, M. and Rostamizadeh, A. , title =. JMLR , volume =. 2012 , pages =

    2012

  174. [182]

    NIPS , pages =

    On kernel target alignment , author =. NIPS , pages =

  175. [183]

    and Fung, G

    Masaeli, M. and Fung, G. and Dy, J. G. , title =. ICML , year =

  176. [184]

    NIPS , pages =

    Nonparametric regression and classification with joint sparsity constraints , author =. NIPS , pages =

  177. [185]

    NIPS , pages =

    Exploring Large Feature Spaces with Hierarchical Multiple Kernel Learning , author =. NIPS , pages =

  178. [186]

    JMLR , volume =

    Minimax-Optimal Rates For Sparse Additive Models Over Kernel Classes Via Convex Programming , author =. JMLR , volume =

  179. [187]

    AISTATS , year =

    Sparse Additive Machine , author =. AISTATS , year =

  180. [188]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =

    Sparse additive models , author =. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =. 2009 , publisher =

    2009

  181. [189]

    NIPS , pages =

    SpAM: Sparse Additive Models , author =. NIPS , pages =

  182. [190]

    Advances in Neural Information Processing Systems 19 , editor =

    Correcting Sample Selection Bias by Unlabeled Data , author =. Advances in Neural Information Processing Systems 19 , editor =

  183. [191]

    , title =

    Roth, V. , title =. IEEE Transactions on Neural Networks , year =

  184. [192]

    and Peng, H

    Ding, C. and Peng, H. , title =. Journal of Bioinformatics an Computaitonal Biology , year =

  185. [193]

    and Andr\'

    Guyon, I. and Andr\'. An introduction to variable and feature selection , journal =. 2003 , volume =

    2003

  186. [194]

    ICML , year =

    From word embeddings to document distances , author =. ICML , year =

  187. [195]

    arXiv preprint arXiv:2104.07710 , year =

    Approximation algorithms for 1-Wasserstein distance between persistence diagrams , author =. arXiv preprint arXiv:2104.07710 , year =

    arXiv

  188. [196]

    ICML , year =

    Linear-complexity data-parallel earth mover’s distance approximations , author =. ICML , year =

  189. [197]

    3rd international workshop on statistical and computational theories of vision , volume =

    Fast image retrieval via embeddings , author =. 3rd international workshop on statistical and computational theories of vision , volume =. 2003 , organization =

    2003

  190. [198]

    CVPR , year =

    Superglue: Learning feature matching with graph neural networks , author =. CVPR , year =

  191. [199]

    ALENEX , year =

    Approximating 1-Wasserstein Distance between Persistence Diagrams by Graph Sparsification , author =. ALENEX , year =

  192. [200]

    Transactions of the Association for Computational Linguistics , volume =

    Back to basics for monolingual alignment: Exploiting word similarity and contextual evidence , author =. Transactions of the Association for Computational Linguistics , volume =. 2014 , publisher =

    2014

  193. [201]

    NeurIPS , year =

    Projection robust Wasserstein distance and Riemannian optimization , author =. NeurIPS , year =

  194. [202]

    ICML , year =

    Wasserstein generative adversarial networks , author =. ICML , year =

  195. [203]

    NeurIPS , year =

    Improved training of wasserstein gans , author =. NeurIPS , year =

  196. [204]

    NeurIPS , year =

    Joint distribution optimal transportation for domain adaptation , author =. NeurIPS , year =

  197. [205]

    ICML , year =

    Fast computation of Wasserstein barycenters , author =. ICML , year =

  198. [206]

    SIAM Journal on Scientific Computing , volume =

    Iterative Bregman projections for regularized transportation problems , author =. SIAM Journal on Scientific Computing , volume =. 2015 , publisher =

    2015

  199. [207]

    AAAI , year =

    Wasserstein distance guided representation learning for domain adaptation , author =. AAAI , year =

  200. [208]

    International Conference on Scale Space and Variational Methods in Computer Vision , pages =

    Wasserstein barycenter and its application to texture mixing , author =. International Conference on Scale Space and Variational Methods in Computer Vision , pages =. 2011 , organization =

    2011

  201. [209]

    NIPS , year =

    Supervised word mover's distance , author =. NIPS , year =

  202. [210]

    Zhao, Wei and Peyrard, Maxime and Liu, Fei and Gao, Yang and Meyer, Christian M and Eger, Steffen , journal =. Mover

  203. [211]

    Journal of machine learning research , volume =

    Learning to match via inverse optimal transport , author =. Journal of machine learning research , volume =

  204. [212]

    NeurIPS , year =

    Differentiable ranking and sorting using optimal transport , author =. NeurIPS , year =

  205. [213]

    ICML , year =

    Fast differentiable sorting and ranking , author =. ICML , year =

  206. [214]

    NeurIPS , year =

    Differentiable top-k with optimal transport , author =. NeurIPS , year =

  207. [215]

    AISTATS , year =

    Learning generative models with sinkhorn divergences , author =. AISTATS , year =

  208. [216]

    EMNLP , year =

    Word Rotator's Distance , author =. EMNLP , year =

  209. [217]

    AISTATS , year =

    Entropy partial transport with tree metrics: Theory and practice , author =. AISTATS , year =

  210. [218]

    ICML , year =

    Re-evaluating Word Mover's Distance , author =. ICML , year =

  211. [219]

    and Suzuki, Taiji

    Tomioka, Ryota. and Suzuki, Taiji. and Sugiyama, Masashi. , title =. JMLR , year =

  212. [220]

    Proceedings 42nd IEEE Symposium on Foundations of Computer Science , pages =

    Algorithmic applications of low-distortion geometric embeddings , author =. Proceedings 42nd IEEE Symposium on Foundations of Computer Science , pages =. 2001 , organization =

    2001

  213. [221]

    ICML , year =

    Scalable nearest neighbor search for optimal transport , author =. ICML , year =

  214. [222]

    , booktitle =

    Seeger, M. , booktitle =. Covariance kernels from

  215. [223]

    NIPS , pages =

    From lasso regression to feature vector machine , author =. NIPS , pages =

  216. [224]

    K. B. Petersen and M. S. Pedersen , title =. 2008 , keywords =

    2008

  217. [225]

    Nonlinear principal component analysis as a kernel eigenvalue problem , journal =

    Sch\". Nonlinear principal component analysis as a kernel eigenvalue problem , journal =. 1998 , pages =

    1998

  218. [226]

    Feuerverger , title =

    A. Feuerverger , title =. International Statistical Review , year =

  219. [227]

    Kankainen , title =

    A. Kankainen , title =. 1995 , address =

    1995

  220. [228]

    Steinwart , title =

    I. Steinwart , title =. JMLR , year =

  221. [229]

    , title =

    Aronszajn, N. , title =. Transactions of the American Mathematical Society , year =

  222. [230]

    NIPS , pages =

    A Kernel Statistical Test of Independence , author =. NIPS , pages =

  223. [231]

    NIPS , year =

    Kernel Measures of Conditional Dependence , author =. NIPS , year =

  224. [232]

    Dinh, T. P. and An, L. H. , title =. SIAM Journal on Optimization , volume =. 1988 , pages =

    1988

  225. [233]

    Pardalos, P. M. and Rendl, F. and Wolkowicz, H. , title =. DIMACS Series in Discrete Mathematics and Theoritical Computer Science , year =

  226. [234]

    and Wright, S

    Nocedal, J. and Wright, S. , title =

  227. [235]

    Quackenbush, S. R. and Barnwell, T. P. and Clements, M. A. , title =

  228. [236]

    and Dell'Amico, M

    Burkard, R. and Dell'Amico, M. and Martello, Silvano. , title =

  229. [237]

    and Russell, B

    Torralba, A. and Russell, B. C. and Yuen, Jenny , title =. Proceedings of the IEEE , volume =. 2010 , pages =

    2010

  230. [238]

    Proceedings of the NIPS 2009 Workshop on Discrete Optimization in Machine Learning: Submodularity, Sparsity and Polyhedra (DISCML) , address =

    Hai Nguyen and Katrin Franke and Slobodan Petrovic , title =. Proceedings of the NIPS 2009 Workshop on Discrete Optimization in Machine Learning: Submodularity, Sparsity and Polyhedra (DISCML) , address =. 2009 , pages =

    2009

  231. [239]

    ICML , year =

    Feature selection for high-dimensional genomic microarray data , author =. ICML , year =

  232. [240]

    , booktitle =

    Forman, G. , booktitle =

  233. [241]

    and Macon, M

    Kain, A. and Macon, M. W. , title =. Proceedings of 1998 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP1998) , address =. 1988 , pages =

    1998

  234. [242]

    and Nakamura, S

    Abe, M. and Nakamura, S. and Shikano, K. and Kuwabara, H. , title =. Proceedings of 1988 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP1988) , address =. 1988 , pages =

    1988

  235. [243]

    Cancer cell , volume =

    Gene expression correlates of clinical prostate cancer behavior , author =. Cancer cell , volume =

  236. [244]

    Minka, T. P. , title =. 2000 , author1_url =

    2000

  237. [245]

    and Suzuki, T

    Kanamori, T. and Suzuki, T. and Sugiyama, M. , title =. 2009 , author1_url =

    2009

  238. [246]

    and Burkard, R

    Finke, G. and Burkard, R. E. and Rendl, F. , title =. Annals of Discrete Mathematics , volume =. 1987 , pages =

    1987

  239. [247]

    , title =

    Kuhn, H.W. , title =. Naval Research Logistics Quarterly , volume =. 1955 , pages =

    1955

  240. [248]

    and Suzuki, T

    Kanamori, T. and Suzuki, T. and Sugiyama, M. , title =. IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences , volume =. 2010 , pages =

    2010

  241. [249]

    and Hido, S

    Kanamori, T. and Hido, S. and Sugiyama, M. , title =. JMLR , volume =. 2009 , pages =

    2009

  242. [250]

    Proceedings of the National Academy of Sciences , volume =

    Statistical learning and selective inference , author =. Proceedings of the National Academy of Sciences , volume =. 2015 , publisher =

    2015

  243. [251]

    Science , year =

    Nonlinear Dimensionality Reduction by Locally Linear Embedding , author =. Science , year =

  244. [252]

    and Fokoue, E

    Gunduz, N. and Fokoue, E. , year =

  245. [253]

    Sriperumbudur, B. K. and Fukumizu, K. and Lanckriet, G. RG. , journal =. Universality, characteristic kernels and

  246. [254]

    , title =

    Jebara, T. , title =. Conference on Computational Learning theory (COLT) , year = 2004, pages =

    2004

  247. [255]

    and Huerta, R

    Rodriguez-Lujan, I. and Huerta, R. and Elkan, C. and Cruz, C. S. , title =. JMLR , volume =. 2010 , pages =

    2010

  248. [256]

    and Kersting, K

    Quadrianto, N. and Kersting, K. and Tuytelaars, T. and Buntine, W. L. , title =. Proceedings of the international conference on Multimedia information retrieval , year = 2010, address =

    2010

  249. [257]

    arXiv preprint arXiv:1506.04364 , year =

    Localized Multiple Kernel Learning---A Convex Approach , author =. arXiv preprint arXiv:1506.04364 , year =

    arXiv

  250. [258]

    ICML , year =

    Localized multiple kernel learning , author =. ICML , year =

  251. [259]

    and Wang, L

    Zhao, Z. and Wang, L. and Li, H , title =

  252. [260]

    and Juarez, S

    Jagarlamudi, J. and Juarez, S. and Daum\'. Kernelized Sorting for Natural Language Processing , booktitle =

  253. [261]

    , journal =

    Tibshirani, R. , journal =. Regression shrinkage and selection via the. 1996 , volume =

    1996

  254. [262]

    The Annals of Statistics , year =

    Kernel dimension reduction in regression , author =. The Annals of Statistics , year =

  255. [263]

    and Heckerman, D

    Geiger, D. and Heckerman, D. , title =. 10th Annual Conference on Uncertainty in Artificial Intelligence (UAI1994) , year = 1994, pages =

    1994

  256. [264]

    and Hoyer, P

    Shimizu, S. and Hoyer, P. O. and Hyv\"arinen, A. and Kerminen, A. J. , journal =. A Linear Non-. 2006 , volume =

    2006

  257. [265]

    Physical Review E , year =

    Estimating Mutual Information , author =. Physical Review E , year =

  258. [266]

    and Bousquet, O

    Gretton, A. and Bousquet, O. and Smola, Alex. and Sch. Measuring statistical dependence with. ALT , year =

  259. [267]

    ICML , pages =

    Supervised feature selection via dependence estimation , author =. ICML , pages =

  260. [268]

    , booktitle =

    Ng, A.Y. , booktitle =. Feature selection,

  261. [269]

    and Janzing, D

    Mooij, J. and Janzing, D. and Peters, J. and Sch\"olkopf, B , title =. ICML , year =

  262. [270]

    IEEE Transactions on Biomedical Engineering , year =

    Online electromyographic control of a robotic prosthesis , author =. IEEE Transactions on Biomedical Engineering , year =

  263. [271]

    The Annals of Statistics , volume =

    Valid post-selection inference , author =. The Annals of Statistics , volume =. 2013 , publisher =

    2013

  264. [272]

    AISTATS , year =

    Sufficient dimension reduction via squared-loss mutual information estimation , author =. AISTATS , year =

  265. [273]

    arXiv preprint arXiv:1706.02262 , year =

    Infovae: Information maximizing variational autoencoders , author =. arXiv preprint arXiv:1706.02262 , year =

    arXiv

  266. [274]

    AAAI , year =

    Dependence minimizing regression with model selection for non-linear causal inference under non-Gaussian noise , author =. AAAI , year =

  267. [275]

    ICML , year =

    Mutual Information Neural Estimation , author =. ICML , year =

  268. [276]

    2009 IEEE International Symposium on Information Theory , pages =

    Mutual information approximation via maximum likelihood estimation of density ratio , author =. 2009 IEEE International Symposium on Information Theory , pages =. 2009 , organization =

    2009

  269. [277]

    BMC Bioinformatics , year =

    Mutual Information Estimation Reveals Global Associations between Stimuli and Biological Processes , author =. BMC Bioinformatics , year =

  270. [278]

    Journal of Machine Learning Research , volume =

    Dimensionality reduction for supervised learning with reproducing kernel Hilbert spaces , author =. Journal of Machine Learning Research , volume =

  271. [279]

    NIPS , year =

    Exact post model selection inference for marginal screening , author =. NIPS , year =

  272. [280]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =

    Large covariance estimation by thresholding principal orthogonal complements , author =. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume =. 2013 , publisher =

    2013

  273. [281]

    and Janzing, D

    Mooij, J. and Janzing, D. and Sch\"olkopf, B , title =

  274. [282]

    Journal of chemical information and modeling , volume =

    Quantitative structure--activity relationship models for ready biodegradability of chemicals , author =. Journal of chemical information and modeling , volume =. 2013 , publisher =

    2013

  275. [283]

    Hoyer, P. O. and Janzing, D. and Mooij, J. M. and Peters, J. and Sch\"olkopf, B. , title =. NIPS , editor =. 2009 , address =

    2009

  276. [284]

    Energy and Buildings , volume =

    Accurate quantitative estimation of energy performance of residential buildings using statistical machine learning tools , author =. Energy and Buildings , volume =. 2012 , publisher =

    2012

  277. [285]

    2015 , publisher =

    Statistical learning with sparsity: the lasso and generalizations , author =. 2015 , publisher =

    2015

  278. [286]

    NIPS , year =

    Sugiyama, Masashi and Nakajima, Shinichi and Kashima, Hisashi and von B\"unau, Paul and Kawanabe, Motoaki , title =. NIPS , year =

  279. [287]

    Sch\"olkopf and A

    B. Sch\"olkopf and A. J. Smola , title =

  280. [288]

    C. R. Shelton , title =

  281. [289]

    Birkenes , title =

    O. Birkenes , title =

  282. [290]

    Tanabe , title =

    K. Tanabe , title =. 2001 , number =

    2001

  283. [291]

    Bishop, C. M. , title =. 2006 , address =

    2006

  284. [292]

    1998 , address =

    Learning to Learn , publisher =. 1998 , address =

    1998

  285. [293]

    A kernel approach for learning from almost orthogonal patterns , booktitle =

    Sch\". A kernel approach for learning from almost orthogonal patterns , booktitle =. 2002 , pages =

    2002

  286. [294]

    Peng and F

    H. Peng and F. Long and C. Ding , title =. IEEE Transactions on Pattern Analysis and Machine Intelligence , year =

  287. [295]

    Quadrianto and A.J

    N. Quadrianto and A.J. Smola and L. Song and T. Tuytelaars , title =. IEEE Transactions on Pattern Analysis and Machine Intelligence , year =

  288. [296]

    Kanamori and T

    T. Kanamori and T. Suzuki and M. Sugiyama , title =

  289. [297]

    2000 , address =

    Causality: Models, Reasoning and Inference , author =. 2000 , address =

    2000

  290. [298]

    Hardy, G. H. and Littlewood, J. E. and P\'. Inequalities , publisher =. 1952 , address =

    1952

  291. [299]

    Patriksson , title =

    M. Patriksson , title =

  292. [300]

    and Vandenberghe, L

    Boyd, S. and Vandenberghe, L. , title =. 2004 , address =

    2004

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.