Pith. sign in

REVIEW 3 major objections 4 minor 66 references

Open-Set Heterogeneous Domain Adaptation: Theoretical Analysis and Algorithm

T0 review · 3 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read A new three-term error bound for open-set heterogeneous domain adaptation shows exactly what a model must optimize, and a method built on that bound outperforms all baselines on 56 tasks.

desk verdict Fresh DA setting with a very solid benchmark, but the central finite-sample bound has a proof error that should not be ignored. read the letter →

arxiv 2412.13036 v1 pith:CSCRQCOZ submitted 2024-12-17 cs.LG

classification cs.LG
keywords open-setheterogeneousdomainadaptationlearningboundsrepresentationpseudo-labelingnon-negativeriskestimatornovelclassdetection
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

This paper defines and analyzes open-set heterogeneous domain adaptation (OSHeDA), the realistic setting where source and target data live in different feature spaces and the target also contains classes never seen in the source, as when a model trained on digital ECG traces must read scanned paper ECGs and catch rare abnormalities. The authors prove an upper bound on target prediction error that splits into three pieces: source error, an open-set difference between how the two domains treat the unknown class, and the distance between source and target representations of known classes. A companion lower bound shows that pushing unknown-class representations away from known ones is necessary to keep target error from growing. Guided by these bounds, they build RL-OSHeDA, which aligns known-class representations with an MMD-based term, segregates unknowns, and uses a non-negative risk estimator with two-stage pseudo-labeling. Across 56 tasks on text, image, and clinical ECG data, the method reports higher harmonic-mean accuracy than every compared baseline.

What carries the argument

The load-bearing object is the error decomposition in Theorem 1, which separates target error into source error, open-set difference, and known-class domain distance. The open-set difference $E(P^u_t,h\circ f_t)-\lambda E(P^u_s,h\circ f_s)$ is the term unique to open-set settings; Proposition 1's lower bound motivates the segregation term by showing that separating the unknown-class target distribution from the source distribution lowers the floor on target error. The finite-sample Theorem 2 adds two pieces of machinery: Natarajan-dimension complexity terms from Lemma 5, and a pseudo-label noise $N(P_{t,k},g)$ that enters the bound through the triangle inequality for JS divergence. On the algorithm side, these terms are implemented with centroid-based MMD distances for $L_{inv}$ and $L_{seg}$, and with the non-negative risk estimator of Eq. (5) for $L_{osd}$. Together the bound and its finite-sample refinement tell the method what to optimize and why.

What would settle it

Run a synthetic OSHeDA task with oracle labels for unknown-class target samples, train a variant of RL-OSHeDA that uses those oracle labels to compute the true open-set difference, and compare its target error against the version using the non-negative risk estimator in Eq. (5); if the two diverge when the unknown-class loss on known samples is nonzero, the finite-sample guarantee's reliance on unavailable empirical errors is exposed.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that target-domain error in OSHeDA obeys a three-term decomposition. With $\lambda=P_t(Y\in\mathcal{Y}_s)$ the share of target samples from known classes and $C$ the loss bound, Theorem 1 gives $E(P_t,h\circ f_t)\le \lambda E(P_s,h\circ f_s)+\bigl(E(P^u_t,h\circ f_t)-\lambda E(P^u_s,h\circ f_s)\bigr)+\sqrt{2\lambda C}\bigl(D_{JS}(P_s(Z)\|P_{t,k}(Z))^{1/2}+D_{JS}(P_s(Z,Y)\|P_{t,k}(Z,Y))^{1/2}\bigr)$. The three terms are the source error, the open-set difference, and the known-class domain distance in representation space. The proposed method, Representation Learning for OSHeDA (RL-OSHeDA), is constructed to minimize exactly these quantities: classification loss on labeled data, MMD alignment of marginal and label-conditioned known-class representations, a segregation term for the unknown class, and a non-negative risk estimator whose optimum is the open-set difference. The finite-sample version of the bound adds a pseudo-label noise term, which the two-stage pseudo-labeling procedure is designed to reduce. The empirical claim is that this combination outperforms prior HeDA, OSDA, and OS-SSL methods on all 56 evaluated tasks.

Load-bearing premise

The finite-sample guarantee depends on being able to measure the classifier's error on the very unknown-class target samples that come unlabeled, and the algorithm's substitute estimator only matches that measure under a condition the setup does not guarantee.

Editorial extensions

If this is right

  • If Theorem 1 is right, a successful OSHeDA method must simultaneously reduce source error, control the open-set difference, and align known-class marginal and joint distributions; dropping any one of these terms should raise target error.
  • Because the joint-distribution term includes the target label distribution, labeled target data or high-quality pseudo-labels are necessary in OSHeDA; unlabeled target data alone cannot minimize the bound.
  • The lower bound implies that separating unknown-class representations from known-class ones is a requirement, not a heuristic, because it directly prevents the lower bound on target error from growing.
  • The finite-sample bound shows pseudo-label noise enters additively, so the two-stage pseudo-labeling scheme is not an engineering convenience but a way to make the bound tighter.
  • The reported consistent ranking advantage over eight baselines, with pairwise significance p-values below 0.05 on 56 tasks, is the empirical consequence predicted by the theory.

Reading between the lines

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

  • A testable extension the paper leaves implicit: the three-term decomposition predicts that RL-OSHeDA's margin over baselines should grow as the unknown-class fraction $1-\lambda$ increases and as feature-space heterogeneity grows, because both enlarge the open-set difference and the JS-distance terms; sweeping $\lambda$ on a fixed benchmark would check this.
  • The theory collapses all unknown classes into one 'unk' label; extending the bound to multiple unknown classes with their own conditional distributions would change the open-set difference term and likely require per-class segregation terms.
  • Equation (5) is the only component that estimates the open-set difference without unknown-class labels; on a synthetic task with oracle unknown labels, comparing its value with the true open-set difference would show how tightly the finite-sample motivation tracks the implemented loss.
  • Because open-set domain adaptation and positive-unlabeled learning are special cases of the same decomposition, the analysis could serve as a template for unifying those neighboring problems.
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

3 major / 4 minor

Summary. The paper formalizes open-set heterogeneous domain adaptation (OSHeDA), in which source and target feature spaces differ and the target label space strictly contains the source label space. It develops infinite-sample upper and lower bounds for the target error (Theorem 1 and Proposition 1), a finite-sample bound with pseudo-label noise (Theorem 2), and a homogeneous-DA comparison (Proposition 3). Guided by these bounds, it proposes RL-OSHeDA, a representation-learning method with two feature mappings, MMD-based alignment and segregation terms, a non-negative open-set risk estimator, and a two-stage pseudo-labeling scheme. The method is evaluated on 56 tasks across seven datasets, where it reports consistent improvements over baselines and releases code.

Significance. The OSHeDA scenario is timely, and the empirical evaluation is broad, spanning clinical, vision, and text data. The theoretical decomposition into source error, open-set difference, and known-class domain distance is a useful organizing principle, and the lower-bound argument for segregating unknown classes is intuitive. The paper also makes a concrete algorithmic contribution and publicly releases code, which strengthens reproducibility. However, the theoretical narrative currently contains a false inequality in the proof of the main finite-sample theorem and a statement-level typo in the proof of Proposition 1; these need repair before the bounds can be cited as proven.

major comments (3)
  1. [Appendix A.2, Eq. (21)] Equation (21) in the proof of Theorem 2 invokes a triangle inequality for Jensen–Shannon divergence, DJS(A||B) ≤ DJS(A||C)+DJS(C||B), which is false. For example, with point masses A=(1,0), B=(0,1), C=(0.8,0.2), DJS(A||B)=0.693 while DJS(A||C)+DJS(C||B)=0.498. The inequality holds for the square root of JS divergence: sqrt(DJS(A||B)) ≤ sqrt(DJS(A||C))+sqrt(DJS(C||B)). Since the final bound in Theorem 2 is exactly of the square-root form, the theorem can be repaired by applying the metric property of sqrt(DJS) in place of the current Eq. (21), but the proof as written is not valid.
  2. [Appendix A.2, Proposition 1 proof] The final displayed inequality in the proof of Proposition 1 reads λ E(Pt,u,h∘ft), whereas the theorem statement and the preceding Eqs. (19)–(20) imply λ E(Pt,k,h∘ft). As written, the proof does not match the proposition; this should be corrected.
  3. [Section 4.2 / Section 5.2] Theorem 2 treats the pseudo-label model g as a fixed object and states a bound for all h with N(Pt,k,g) appearing in the bound. In the algorithm of Section 5.2, g is re-derived from the current h and the unlabeled target data at every stage-2 update. The theorem therefore does not, as stated, cover the adaptively coupled (g,h) used by RL-OSHeDA. The authors should either restrict the theoretical claim to a fixed g and state that the algorithm is an approximation, or provide a fixed-point argument.
minor comments (4)
  1. [Appendix A.1, Lemma 5 proof] In the proof of Lemma 5, the text says 'because of Assumption 2' when bounding the squared losses by C^2; this should cite Assumption 1 instead.
  2. [Table 1] The KPG row reports N/A for PTB-XL but KPG is included in the average row; the authors should state explicitly that the average excludes the N/A entry.
  3. [Tables 12 and 19] The dataset names in these tables use the typo 'Reset-50' for ResNet-50; this should be corrected throughout.
  4. [Section 4.1] The notation P^u_t is easy to confuse with Pt,u; the authors should add a one-sentence reminder in Section 4.2 when empirical errors are introduced, since P^u_t is the surrogate-unk distribution, not the unknown-class conditional.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity; the learning bounds are derived from independent JS-divergence and Rademacher arguments, and the self-citations are peripheral rather than load-bearing.

full rationale

The central derivation chain is self-contained. Theorem 1 is obtained by decomposing the target risk as λE(Pt,k,h∘ft)+(1−λ)E(Pt,u,h∘ft) (Eq. 16), then applying Lemma 1 to the known-class pair and Lemma 2 to the induced unknown-label distributions; the resulting bound contains a source-error term, an open-set difference, and JS-divergence domain distances. Theorem 2 adds a finite-sample Rademacher analysis (Lemmas 4 and 5) and a pseudo-label noise term N(Pt,k,g); this is an upper bound stated in terms of the model's own pseudo-label quality, not a fitted parameter renamed as a prediction. The algorithm's objective L=Lcls+Linv−Lseg+Losd is explicitly motivated by the bound, and Losd in Eq. (5) is exactly an empirical, non-negative version of the open-set difference E(Pu_t,h∘ft)−λE(Pu_s,h∘fs) from Theorem 2; this is theory-to-algorithm design, not circularity. The self-citations to Pham, Zhang, and Zhang (2023) for representation learning and conditional alignment, and to Pham, Zhang, and Zhang (2024) for making cross-entropy bounded under Assumption 1, are not load-bearing: the theorems rest on standard divergence inequalities and Rademacher bounds, and the cited works provide background techniques rather than the target result. The reader's concern about needing unknown-class labels misreads P^u_t: by definition f^u_t(X_t,Y)=(X_t,unk), so bE(P^u_t,h∘ft) is computable from all unlabeled target data; Eq. (5) is a valid empirical estimate of the open-set difference. The separate concern that the proof of Theorem 2 uses a triangle inequality for DJS is a mathematical-correctness issue, not a circular derivation, and therefore does not affect the circularity score. Overall, no claimed prediction reduces by construction to its own input, and no load-bearing argument depends on an unverified self-citation chain.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the bounded-loss assumption (Assumption 1) and the availability of labeled known-class target data. The algorithm introduces lambda (estimated via PU learning) and stage threshold T (not reported), but no new entities. Lemma 3's proof uses an unstated bounded density ratio, a hidden assumption in a supporting result.

free parameters (2)
  • lambda (known class prior) = estimated per dataset via PU learning (Zeiberg et al. 2020)
    Weights source and target losses and determines the 1-lambda fraction of unlabeled target assigned to the unknown class; affects which samples receive unknown pseudo-labels.
  • T (stage-switch threshold)
    Controls when training switches from Lcls alone to the full objective with pseudo-labels; no value or selection procedure is reported in the paper.
assumptions (4)
  • standard math Loss function L is upper bounded by a constant C (Assumption 1)
    Used in all proofs (Lemma 1, Theorems 1-2); the paper notes cross-entropy must be adjusted to satisfy it.
  • domain assumption Sufficient representation: Is(Z,Y)=Is(Xs,Y) (Assumption 2)
    Used only in Proposition 3 (HoDA bound) to show covariate shift is preserved in representation space.
  • domain assumption Shared representation space Z and a single classifier h over Z work for both domains
    Core to the representation learning approach; not proven, standard in HeDA.
  • ad hoc to paper Bounded density ratio pt(x,y)/ps(x,y) in Lemma 3's proof
    The proof of Lemma 3 uses max_{x,y} pt/ps as a finite constant, but boundedness is not stated as an assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Open-Set Heterogeneous Domain Adaptation: Theoretical Analysis and Algorithm." pith.science (2026). https://pith.science/paper/CSCRQCOZ

@misc{pith2026241213036,
  author       = {Pith},
  title        = {Pith review of: Open-Set Heterogeneous Domain Adaptation: Theoretical Analysis and Algorithm},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CSCRQCOZ}},
  note         = {Machine review of arXiv:2412.13036}
}
read the original abstract

Domain adaptation (DA) tackles the issue of distribution shift by learning a model from a source domain that generalizes to a target domain. However, most existing DA methods are designed for scenarios where the source and target domain data lie within the same feature space, which limits their applicability in real-world situations. Recently, heterogeneous DA (HeDA) methods have been introduced to address the challenges posed by heterogeneous feature space between source and target domains. Despite their successes, current HeDA techniques fall short when there is a mismatch in both feature and label spaces. To address this, this paper explores a new DA scenario called open-set HeDA (OSHeDA). In OSHeDA, the model must not only handle heterogeneity in feature space but also identify samples belonging to novel classes. To tackle this challenge, we first develop a novel theoretical framework that constructs learning bounds for prediction error on target domain. Guided by this framework, we propose a new DA method called Representation Learning for OSHeDA (RL-OSHeDA). This method is designed to simultaneously transfer knowledge between heterogeneous data sources and identify novel classes. Experiments across text, image, and clinical data demonstrate the effectiveness of our algorithm. Model implementation is available at \url{https://github.com/pth1993/OSHeDA}.

Figures

Figures reproduced from arXiv: 2412.13036 by the authors.

Figure 1
Figure 1. A motivating example about OSHeDA in the con [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overall architecture of RL-OSHeDA is illustrated with a motivating example from ECG-based diagnosis application. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Critical Difference diagram for all methods calcu [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Performances w.r.t. different number of labeled tar [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

66 extracted references · 47 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    H.; and Mitliagkas, I

    Albuquerque, I.; Monteiro, J.; Darvishi, M.; Falk, T. H.; and Mitliagkas, I. 2019. Generalizing to unseen domains via distribution matching. arXiv preprint arXiv:1911.00804

  4. [4]

    R.; Usunier, N.; and Goutte, C

    Amini, M. R.; Usunier, N.; and Goutte, C. 2009. Learning from multiple partially observed views-an application to multilingual text categorization. Advances in neural information processing systems, 22

  5. [5]

    Ben-David, S.; Blitzer, J.; Crammer, K.; Kulesza, A.; Pereira, F.; and Vaughan, J. W. 2010. A theory of learning from different domains. Machine learning, 79: 151--175

  6. [6]

    Biau, G.; Cadre, B.; Sangnier, M.; and Tanielian, U. 2020. Some theoretical properties of GANS . The Annals of Statistics, 48(3): 1539 -- 1566

  7. [7]

    R.; and Tommasi, T

    Bucci, S.; Loghmani, M. R.; and Tommasi, T. 2020. On the effectiveness of image rotation for open set domain adaptation. In European conference on computer vision, 422--438. Springer

  8. [8]

    Chen, Y.; Zhu, X.; Li, W.; and Gong, S. 2020. Semi-supervised learning under class distribution mismatch. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 34, 3569--3576

Show all 66 references
  1. [9]

    Chua, T.-S.; Tang, J.; Hong, R.; Li, H.; Luo, Z.; and Zheng, Y. 2009. Nus-wide: a real-world web image database from national university of singapore. In Proceedings of the ACM international conference on image and video retrieval, 1--9

  2. [10]

    Cortes, C.; and Mohri, M. 2014. Domain adaptation and sample bias correction theory and algorithm for regression. Theoretical Computer Science, 519: 103--126

  3. [11]

    Dem s ar, J. 2006. Statistical comparisons of classifiers over multiple data sets. The Journal of Machine learning research, 7: 1--30

  4. [12]

    Deng, J.; Dong, W.; Socher, R.; Li, L.-J.; Li, K.; and Fei-Fei, L. 2009. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, 248--255. Ieee

  5. [13]

    Donahue, J.; Jia, Y.; Vinyals, O.; Hoffman, J.; Zhang, N.; Tzeng, E.; and Darrell, T. 2014. Decaf: A deep convolutional activation feature for generic visual recognition. In International conference on machine learning, 647--655. PMLR

  6. [14]

    Du, P.; Zhao, S.; Sheng, Z.; Li, C.; and Chen, H. 2023. Semi-Supervised Learning via Weight-aware Distillation under Class Distribution Mismatch. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 16410--16420

  7. [15]

    Fang, Z.; Lu, J.; Liu, F.; Xuan, J.; and Zhang, G. 2020. Open set domain adaptation: Theoretical bound and algorithm. IEEE transactions on neural networks and learning systems, 32(10): 4309--4322

  8. [16]

    Fang, Z.; Lu, J.; Liu, F.; and Zhang, G. 2022. Semi-supervised heterogeneous domain adaptation: Theory and algorithms. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(1): 1087--1105

  9. [17]

    Ganin, Y.; Ustinova, E.; Ajakan, H.; Germain, P.; Larochelle, H.; Laviolette, F.; Marchand, M.; and Lempitsky, V. 2016. Domain-adversarial training of neural networks. The journal of machine learning research, 17(1): 2096--2030

  10. [18]

    Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; and Bengio, Y. 2014. Generative adversarial nets. Advances in neural information processing systems, 27

  11. [19]

    N.; Patel, J.; Kochendorfer, K

    Grief, S. N.; Patel, J.; Kochendorfer, K. M.; Green, L. A.; Lussier, Y. A.; Li, J.; Burton, M.; and Boyd, A. D. 2016. Simulation of ICD-9 to ICD-10-CM transition for family medicine: simple or convoluted? The Journal of the American Board of Family Medicine, 29(1): 29--36

  12. [20]

    Griffin, G.; Holub, A.; Perona, P.; et al. 2007. Caltech-256 object category dataset. Technical report, Technical Report 7694, California Institute of Technology Pasadena

  13. [21]

    Gu, X.; Yang, Y.; Zeng, W.; Sun, J.; and Xu, Z. 2022. Keypoint-guided optimal transport with applications in heterogeneous domain adaptation. Advances in Neural Information Processing Systems, 35: 14972--14985

  14. [22]

    Guo, L.-Z.; Zhang, Z.-Y.; Jiang, Y.; Li, Y.-F.; and Zhou, Z.-H. 2020. Safe deep semi-supervised learning for unseen-class unlabeled data. In International conference on machine learning, 3897--3906. PMLR

  15. [23]

    He, R.; Han, Z.; Lu, X.; and Yin, Y. 2022. Safe-student for safe deep semi-supervised learning with unseen-class unlabeled data. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 14585--14594

  16. [24]

    Hoffman, J.; Rodner, E.; Donahue, J.; Darrell, T.; and Saenko, K. 2013. Efficient learning of domain-invariant image representations. arXiv preprint arXiv:1301.3224

  17. [25]

    Hoffman, J.; Rodner, E.; Donahue, J.; Kulis, B.; and Saenko, K. 2014. Asymmetric and category invariant feature transformations for domain adaptation. International journal of computer vision, 109: 28--41

  18. [26]

    Huang, J.; Fang, C.; Chen, W.; Chai, Z.; Wei, X.; Wei, P.; Lin, L.; and Li, G. 2021. Trash to treasure: Harvesting ood data with cross-modal matching for open-set semi-supervised learning. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 8310--8319

  19. [27]

    Huang, Z.; Yang, J.; and Gong, C. 2022. They are not completely useless: Towards recycling transferable unlabeled data for class-mismatched semi-supervised learning. IEEE Transactions on Multimedia, 25: 1844--1857

  20. [28]

    C.; and Sugiyama, M

    Kiryo, R.; Niu, G.; Du Plessis, M. C.; and Sugiyama, M. 2017. Positive-unlabeled learning with non-negative risk estimator. Advances in neural information processing systems, 30

  21. [29]

    Kolesnikov, A.; Beyer, L.; Zhai, X.; Puigcerver, J.; Yung, J.; Gelly, S.; and Houlsby, N. 2019. Big transfer (BiT): General visual representation learning. arXiv 2020. arXiv preprint arXiv:1912.11370

  22. [30]

    Koltchinskii, V.; and Panchenko, D. 2000. Rademacher processes and bounding the risk of function learning. In High dimensional probability II, 443--457. Springer

  23. [31]

    Krizhevsky, A. 2009. Learning Multiple Layers of Features from Tiny Images. Master's thesis, University of Tront

  24. [32]

    LeCun, Y.; Bengio, Y.; and Hinton, G. 2015. Deep learning. nature, 521(7553): 436--444

  25. [33]

    Li, G.; Kang, G.; Zhu, Y.; Wei, Y.; and Yang, Y. 2021. Domain consensus clustering for universal domain adaptation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 9757--9766

  26. [34]

    Li, J.; Lu, K.; Huang, Z.; Zhu, L.; and Shen, H. T. 2018. Heterogeneous domain adaptation through progressive alignment. IEEE transactions on neural networks and learning systems, 30(5): 1381--1391

  27. [35]

    H.; and Ding, Z

    Li, S.; Xie, B.; Wu, J.; Zhao, Y.; Liu, C. H.; and Ding, Z. 2020. Simultaneous semantic alignment network for heterogeneous domain adaptation. In Proceedings of the 28th ACM international conference on multimedia, 3866--3874

  28. [36]

    Li, W.; Duan, L.; Xu, D.; and Tsang, I. W. 2013. Learning with augmented features for supervised and semi-supervised heterogeneous domain adaptation. IEEE Transactions on Pattern analysis and machine intelligence, 36(6): 1134--1148

  29. [37]

    Liang, P. 2016. CS229T/STAT231: Statistical learning theory (Winter 2016)

  30. [38]

    Liu, F.; Zhang, G.; and Lu, J. 2020. Heterogeneous domain adaptation: An unsupervised approach. IEEE transactions on neural networks and learning systems, 31(12): 5588--5602

  31. [39]

    Liu, H.; Cao, Z.; Long, M.; Wang, J.; and Yang, Q. 2019. Separate to adapt: Open set domain adaptation via progressive separation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2927--2936

  32. [40]

    Luo, Y.; Wang, Z.; Huang, Z.; and Baktashmotlagh, M. 2020. Progressive graph learning for open-set domain adaptation. In International Conference on Machine Learning, 6468--6478. PMLR

  33. [41]

    Mansour, Y.; Mohri, M.; and Rostamizadeh, A. 2009. Domain adaptation: Learning bounds and algorithms. arXiv preprint arXiv:0902.3430

  34. [42]

    Natarajan, B. K. 1989. On learning sets and functions. Machine Learning, 4: 67--97

  35. [43]

    T.; Tran, T.; Gal, Y.; Torr, P.; and Baydin, A

    Nguyen, A. T.; Tran, T.; Gal, Y.; Torr, P.; and Baydin, A. G. 2021. KL Guided Domain Adaptation. In International Conference on Learning Representations

  36. [44]

    Pham, T.-H.; Zhang, X.; and Zhang, P. 2023. Fairness and Accuracy under Domain Generalization. In The Eleventh International Conference on Learning Representations

  37. [45]

    Pham, T.-H.; Zhang, X.; and Zhang, P. 2024. Non-stationary Domain Generalization: Theory and Algorithm. In The 40th Conference on Uncertainty in Artificial Intelligence

  38. [46]

    Qui \ n onero-Candela, J.; Sugiyama, M.; Schwaighofer, A.; and Lawrence, N. D. 2022. Dataset shift in machine learning. Mit Press

  39. [47]

    R.; Levy, R.; and Vasconcelos, N

    Rasiwasia, N.; Costa Pereira, J.; Coviello, E.; Doyle, G.; Lanckriet, G. R.; Levy, R.; and Vasconcelos, N. 2010. A new approach to cross-modal multimedia retrieval. In Proceedings of the 18th ACM international conference on Multimedia, 251--260

  40. [48]

    Russakovsky, O.; Deng, J.; Su, H.; Krause, J.; Satheesh, S.; Ma, S.; Huang, Z.; Karpathy, A.; Khosla, A.; Bernstein, M.; et al. 2015. Imagenet large scale visual recognition challenge. International journal of computer vision, 115: 211--252

  41. [49]

    Saenko, K.; Kulis, B.; Fritz, M.; and Darrell, T. 2010. Adapting visual category models to new domains. In Computer Vision--ECCV 2010: 11th European Conference on Computer Vision, Heraklion, Crete, Greece, September 5-11, 2010, Proceedings, Part IV 11, 213--226. Springer

  42. [50]

    Saito, K.; Kim, D.; and Saenko, K. 2021. Openmatch: Open-set semi-supervised learning with open-set consistency regularization. Advances in Neural Information Processing Systems, 34: 25956--25967

  43. [51]

    Saito, K.; Yamamoto, S.; Ushiku, Y.; and Harada, T. 2018. Open set domain adaptation by backpropagation. In Proceedings of the European conference on computer vision (ECCV), 153--168

  44. [52]

    Shalev-Shwartz, S.; and Ben-David, S. 2014. Understanding machine learning: From theory to algorithms. Cambridge university press

  45. [53]

    Shen, C.; and Guo, Y. 2018. Unsupervised heterogeneous domain adaptation with sparse feature transformation. In Asian conference on machine learning, 375--390. PMLR

  46. [54]

    Simonyan, K.; and Zisserman, A. 2014. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556

  47. [55]

    Van der Maaten, L.; and Hinton, G. 2008. Visualizing data using t-SNE. Journal of machine learning research, 9(11)

  48. [56]

    I.; Samek, W.; and Schaeffter, T

    Wagner, P.; Strodthoff, N.; Bousseljot, R.-D.; Kreiseler, D.; Lunze, F. I.; Samek, W.; and Schaeffter, T. 2020. PTB-XL, a large publicly available electrocardiography dataset. Scientific data, 7(1): 1--15

  49. [57]

    Wang, Y.; Qiao, P.; Liu, C.; Song, G.; Zheng, X.; and Chen, J. 2023. Out-of-distributed semantic pruning for robust semi-supervised learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 23849--23858

  50. [58]

    Xu, R.; Liu, P.; Zhang, Y.; Cai, F.; Wang, J.; Liang, S.; Ying, H.; and Yin, J. 2020. Joint Partial Optimal Transport for Open Set Domain Adaptation. In IJCAI, 2540--2546

  51. [59]

    Yao, Y.; Zhang, Y.; Li, X.; and Ye, Y. 2019. Heterogeneous domain adaptation via soft transfer network. In Proceedings of the 27th ACM MM, 1578--1586

  52. [60]

    Yao, Y.; Zhang, Y.; Li, X.; and Ye, Y. 2020. Discriminative distribution alignment: A unified framework for heterogeneous domain adaptation. Pattern Recognition, 101: 107165

  53. [61]

    Yu, Q.; Ikami, D.; Irie, G.; and Aizawa, K. 2020. Multi-task curriculum framework for open-set semi-supervised learning. In Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part XII 16, 438--454. Springer

  54. [62]

    A.; Ainslie, J.; Alberti, C.; Ontanon, S.; Pham, P.; Ravula, A.; Wang, Q.; Yang, L.; et al

    Zaheer, M.; Guruganesh, G.; Dubey, K. A.; Ainslie, J.; Alberti, C.; Ontanon, S.; Pham, P.; Ravula, A.; Wang, Q.; Yang, L.; et al. 2020. Big bird: Transformers for longer sequences. Advances in neural information processing systems, 33: 17283--17297

  55. [63]

    Zeiberg, D.; Jain, S.; and Radivojac, P. 2020. Fast nonparametric estimation of class proportions in the positive-unlabeled classification setting. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 34, 6729--6736

  56. [64]

    T.; Zhang, K.; and Gordon, G

    Zhao, H.; Des Combes, R. T.; Zhang, K.; and Gordon, G. 2019. On learning invariant representations for domain adaptation. In International conference on machine learning, 7523--7532. PMLR

  57. [65]

    H.; Cao, W.; Wang, X.; and Tian, S

    Zhao, Y.; Li, S.; Zhang, R.; Liu, C. H.; Cao, W.; Wang, X.; and Tian, S. 2022. Semantic correlation transfer for heterogeneous domain adaptation. IEEE Transactions on Neural Networks and Learning Systems

  58. [66]

    Zou, Y.; Yu, Z.; Kumar, B.; and Wang, J. 2018. Unsupervised domain adaptation for semantic segmentation via class-balanced self-training. In Proceedings of the European conference on computer vision (ECCV), 289--305

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.