Pith. sign in

REVIEW 4 major objections 5 minor 60 references

Multi-Label Bayesian Active Learning with Inter-Label Relationships

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Active learning that reads label links beats five baselines

desk verdict A plausible new combination for multi-label active learning that deserves peer review, but its 'consistently superior' claim is statistically unproven without per-seed analysis. read the letter →

arxiv 2411.17941 v3 pith:FVTDAJCF submitted 2024-11-26 cs.LG

Yuanyuan Qi , Jueqing Lu , Xiaohao Yang , Joanne Enticott , Lan Du This is my paper · ORCID
classification cs.LG
keywords multi-labelactivelearninglabelcorrelationnegativebetascoringrulesensemblepseudolabelingdataimbalance
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 proposes CRAB, a multi-label active learning strategy that uses progressively updated positive and negative label-correlation matrices to guide which unlabeled documents to query for annotation. The central claim is that incorporating inter-label relationships—both co-occurrence and mutual exclusivity—into a beta-scoring-rule acquisition function yields more reliable and higher micro-F1 performance than established multi-label active learning methods. The authors argue this matters because real-world multi-label datasets are large, imbalanced, and have complex label dependencies, so a query strategy that exploits those dependencies can reduce annotation cost while maintaining accuracy.

What carries the argument

The core mechanism is the pair of correlation matrices defined in Eqs. (9) and (10): A(m,n) = P(y_m=1|y_n=1), the empirical probability that label m co-occurs with label n, and NegA(m,n) = P(y_m=-1|y_n=1), the empirical probability that label m is absent given label n is present. These matrices are recomputed after each acquisition iteration from the labeled pool, and they drive both the refined subset selection (label-wise, negative-correlated, hard-to-learn) and the correlation-aware beta score in Eq. (13), where each label's beta score is weighted by a normalized attention coefficient derived from A. This lets the acquisition function treat labels as interrelated rather than isolated, capturing asymmetric and hierarchical dependencies.

What would settle it

Run CRAB on RCV1 with the correlation matrices replaced by random matrices (or by identity matrices) each iteration, keeping the rest of the pipeline unchanged. If micro-F1 does not drop substantially relative to the true matrices, the correlation component is not the source of the claimed advantage; alternatively, if performance is highly sensitive to a single early random seed when the labeled pool is 100, the matrix estimates may be too noisy.

Watch

Extended reading notes

Core claim

The paper's central discovery is that a multi-label active learning acquisition function can be made correlation-aware without training additional correlation models. CRAB maintains two dynamic matrices: a positive co-occurrence matrix A and a negative anti-correlation matrix NegA, each computed from the currently labeled pool via conditional frequency counts. These matrices feed three correlated sampling heuristics—label-wise sampling that removes asymmetrically dependent labels, negative-correlated sampling that targets mutually exclusive label pairs, and hard-to-learn sampling—and also weight the beta scoring rule through an attention-like term. The authors report that this strategy consistently outperforms five baselines, including random sampling, MMC, AUDI, ADAPTIVE, BESRA, and CMAL, across four benchmark multi-label text datasets and synthetic variants with varying imbalance ratios.

Load-bearing premise

The correlation matrices are computed from the currently labeled pool using plain frequency counts, and the paper assumes these are reliable guides for sample selection even when the pool starts at just 100 documents, including a dataset with 102 labels where most conditional counts will be zero or near-zero, and no smoothing is applied.

Editorial extensions

If this is right

  • If CRAB's reported gains hold, multi-label active learning can leverage label correlation at negligible extra computational cost, since the matrices are simple frequency counts.
  • The explicit modeling of negative correlations offers a principled way to resolve mutually exclusive label predictions, which should help in domains with strict label hierarchies, such as legal or medical coding.
  • Because the beta scoring rules are tunable, the framework can be adjusted to different imbalance regimes, potentially extending beyond the four text benchmarks to image or sensor data.
  • The refined-pool sampling stage reduces the number of candidates scored by the acquisition function, which could make large-pool active learning more tractable.

Reading between the lines

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

  • A natural extension would be to quantify uncertainty in the correlation matrix estimates (e.g., with Dirichlet smoothing or Laplace smoothing) early in the learning process, when the labeled pool is only 100 documents and RCV1 has 102 labels; the paper does not test this sensitivity.
  • The same correlation-aware attention weighting could be plugged into other acquisition functions beyond beta scoring, such as entropy or margin-based learners, to test whether the benefit is tied to the specific scoring rule.
  • One could design a falsification experiment: shuffle the rows/columns of A and NegA after each iteration while keeping the rest of CRAB fixed. If performance does not degrade significantly relative to the true matrices, the correlation component would not be the source of the reported gains.
  • The paper implicitly assumes label correlations are stable across the unlabeled distribution, but in temporally drifting datasets (e.g., evolving legal text), recomputing matrices from only the labeled pool may not capture shifts; a streaming or exponential-decay update would be a natural test.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes CRAB, a multi-label active learning strategy that maintains dynamically updated positive and negative correlation matrices over the label space, uses ensemble pseudo-labels to refine the unlabeled pool into label-wise, negatively correlated, and hard-to-learn subsets, and then scores the refined pool with a beta scoring rule augmented by an attention-style correlation term. The authors claim that CRAB consistently outperforms several established multi-label active learning baselines on four benchmark datasets across three backbone classifiers, and that it remains robust on synthetic datasets with varying imbalance ratios. The paper also includes ablations for the correlation-aware sampling components, an analysis of hard-to-learn query sizes and decay schedules, and a comparison of clustering-based versus random refinement of the unlabeled pool.

Significance. If the central empirical claim were fully supported, CRAB would be a useful contribution to multi-label active learning: it explicitly models negative and asymmetric label correlations without training additional correlation models, and it combines uncertainty, diversity, and imbalance-aware scoring in a single acquisition function. The paper has several strengths: the code is publicly released, the experimental comparison spans four datasets, three backbones, and five baselines, and the ablation study isolates several components of the method. However, the headline claim that CRAB 'consistently achieves more reliable and superior performance' is not yet supported by the reported evidence, because the experiments lack error bars, confidence intervals, and significance tests, and several key hyperparameters are either inherited from prior work or chosen per dataset without a corresponding sensitivity analysis. The contribution is therefore promising but needs additional empirical validation before the comparative claim can be accepted.

major comments (4)
  1. [Section 4.3, Figures 2–5] The central claim of the paper is comparative: CRAB 'consistently achieves more reliable and superior performance' against established methods (Abstract; Section 4.3). The only quantitative support is the averaged micro-F1 curves in Figures 2–5, whose captions state 'averaged results with 5 random seeds' but report no standard deviation, confidence interval, or significance test. With only five seeds, the reported gaps between CRAB and BESRA or AUDI may be within seed-level noise. The authors should provide per-seed paired comparisons, confidence intervals, or a statistical test (e.g., paired bootstrap or Wilcoxon signed-rank test) across the learning curves or at final acquisition budgets. Without this, the 'consistently superior' claim is underdetermined.
  2. [Section 3.2, Eqs. (9)–(10)] The positive and negative correlation matrices are estimated from the labeled pool by simple conditional frequency counts. The labeled pool starts at only 100 documents (Section 4.1), while RCV1 has 102 labels; many entries of A(m,n) and NegA(m,n) will be estimated from zero or very few observations, and entries with zero denominator are undefined. These matrices are used directly in all three sampling heuristics (Sections 3.3.1–3.3.3) and in the attention score in Eq. (13). The paper provides no smoothing, shrinkage, or uncertainty quantification for these estimates, nor any sensitivity analysis for early active-learning iterations. The authors should analyze the noise in these matrices and demonstrate that the method's behavior is not driven by unreliable early estimates.
  3. [Sections 3.3–3.4 and 4.1] Several decisive hyperparameters are fixed per dataset or inherited from prior work, and the reported sensitivity analysis is not sufficient to establish that the observed gains come from the proposed mechanism. These include the hard-to-learn query size Z (300 for benchmark, 200 for synthetic), the per-label query size N (50 for RCV1, 100 for others), the attention normalization gamma=2 in Eq. (14), the correlation thresholds sigma and 2sigma in Sections 3.3.1–3.3.2, and alpha=0.1, beta=3 from BESRA. Figure 8 only studies the hard-to-learn size and decay on a single configuration, and no sensitivity analysis is reported for gamma, sigma, alpha, beta, or the per-label query size across all datasets. The authors should add a sensitivity study and clarify which choices, if any, were made after seeing evaluation-set performance.
  4. [Sections 3.1 and 3.4, Eq. (13)] The theoretical framework in Section 3.1, culminating in Eqs. (3)–(5), describes a Bayesian expected increase in a proper scoring rule, but the actual acquisition score used in Algorithm 1 is the attention-style beta score SAB in Eq. (13). The connection between the two is not established: Eq. (13) is introduced as a heuristic combination of per-label beta scores weighted by the positive correlation matrix, and it is not shown to approximate or upper-bound the expected score reduction in Eq. (3). The authors should either derive Eq. (13) from the expected-loss-reduction formulation under an explicit approximation, or state clearly that Eq. (13) is a heuristic acquisition score and adjust the claims accordingly.
minor comments (5)
  1. [Section 4.1] There is a typo in 'exployed' (should be 'employed'), and in Section 4.3 'generaliztion' should be 'generalization'.
  2. [Figure 2 caption] The caption 'Averaged micro-F1 score on DistilBERT' is awkward; it should be 'Averaged micro-F1 score using DistilBERT' or 'with DistilBERT'.
  3. [Section 3.3.1 and Eq. (14)] The definition of the threshold sigma as 'the standard deviation of a two-tailed normal distribution' is vague, and Eq. (14) normalizes by the column maximum max(A(:, n)) while the text loosely refers to preserving each label's original significance; the normalization convention should be stated precisely.
  4. [Algorithm 1] The relation between the query size N, the per-label query size N, and the hard-to-learn query size Z is not fully specified: line 12 acquires only N centers after selecting N label-wise and N negative-correlated samples per label plus Z hard samples, and it is unclear how these quantities interact when the per-label sampling produces more than N candidates.
  5. [Section 3.1, Eqs. (4)–(5)] The notation switches between a continuous parameter space theta and a finite ensemble Theta_E without explaining how P(theta | L) is represented in the ensemble approximation; a brief clarification would help readers follow the derivation.

Circularity Check

0 steps flagged · score 2.0 of 10

No load-bearing circularity: CRAB is an empirically evaluated heuristic; the only self-citations (BESRA's alpha/beta and a related-work reference) are not used to derive the central result.

full rationale

CRAB does not present a formal derivation that reduces to its inputs. The correlation matrices A and NegA are computed from the labeled pool and then used in sampling and scoring heuristics (Eqs. 9-14); this is the algorithm's design, not a circular prediction of the same quantity. The beta-scoring parameters alpha=0.1, beta=3 are imported from BESRA (Tan et al. 2024, which shares co-author Lan Du), but they are used as an explicit design choice, not as evidence for CRAB's superiority; CRAB is compared empirically against BESRA and other baselines on four real-world datasets. The related-work citation to Lu et al. 2020 (also co-authored by Jueqing Lu and Lan Du) is descriptive and not load-bearing. No fitted parameter is renamed as a prediction, no uniqueness theorem is invoked, and no derivation step equals its input by construction. The absence of error bars and significance tests in Figs. 2-5 is a statistical-robustness concern, not a circularity concern. Score 2 reflects the presence of minor non-load-bearing self-citations; the central comparative claim stands on independent experiments.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The central claim rests on several hand-chosen hyperparameters and on the reliability of correlation estimates from small labeled pools. None of the free parameters are derived from first principles; most are tuned on the same datasets used for evaluation. No new entities are introduced.

free parameters (7)
  • Beta scoring rule exponents alpha and beta = alpha=0.1, beta=3
    Taken from the greedy search of BESRA (Tan et al., 2024), a paper by the same group. These shape the scoring rule used in every acquisition and imbalance handling, and were selected on the same type of evaluation datasets.
  • Attention normalization factor gamma = 2
    Set by hand in Eq. (14) to 'prevent over-estimating correlated uncertainty'; no principled selection or sensitivity analysis is provided.
  • Asymmetric correlation threshold sigma = described as the standard deviation of a two-tailed normal distribution, effectively 1
    Used in Section 3.3.1 to decide when a label pair is asymmetrically correlated; the definition is vague and no search over sigma is reported.
  • Negative correlation threshold = 2sigma
    Used in Section 3.3.2 to flag mutually exclusive labels; derived from sigma without empirical justification.
  • Hard-to-learn query size Z = 300 (benchmarks), 200 (synthetic)
    Set per dataset in Section 4.1 and shown in Figure 8a to materially affect performance; Section 4.4 states it is 'adjustable across different datasets', implying tuning on the evaluation sets.
  • Per-label query size N = 50 (RCV1), 100 (others)
    Chosen based on label space size in Section 4.1; no sensitivity analysis is provided.
  • Polynomial decay schedule for Z = polynomial (vs. linear/cosine)
    Selected after comparing three schedules in Figure 8b on the evaluation datasets.
assumptions (4)
  • domain assumption The Bayesian expected loss reduction framework (Eqs. 2-5) is a valid model for active learning information gain.
    The paper assumes this framework from Roy and McCallum and Gneiting and Raftery, but the final acquisition function diverges from the exact expected-loss reduction computation.
  • domain assumption The empirical conditional probabilities in A and NegA are accurate estimates of label dependencies.
    Invoked in Section 3.2; the matrices are computed from the current labeled pool, which is very small early in training, with no smoothing or confidence intervals.
  • domain assumption Pseudo labels obtained by thresholding ensemble-averaged probabilities at 0.5 are reliable indicators of hard and conflicting samples.
    Used throughout Section 3.3 to define label-wise, negatively correlated, and hard-to-learn samples; early in active learning the model is poorly trained, so these pseudo labels are noisy.
  • ad hoc to paper The attention-style combination of label scores in Eq. (13) with row-wise normalization in Eq. (14) is a meaningful way to aggregate label informativeness.
    No derivation or comparison to alternative aggregation schemes is provided; gamma=2 and the normalization form are chosen by hand.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multi-Label Bayesian Active Learning with Inter-Label Relationships." pith.science (2026). https://pith.science/paper/FVTDAJCF

@misc{pith2026241117941,
  author       = {Pith},
  title        = {Pith review of: Multi-Label Bayesian Active Learning with Inter-Label Relationships},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FVTDAJCF}},
  note         = {Machine review of arXiv:2411.17941}
}
read the original abstract

The primary challenge of multi-label active learning, differing it from multi-class active learning, lies in assessing the informativeness of an indefinite number of labels while also accounting for the inherited label correlation. Existing studies either require substantial computational resources to leverage correlations or fail to fully explore label dependencies. Additionally, real-world scenarios often require addressing intrinsic biases stemming from imbalanced data distributions. In this paper, we propose a new multi-label active learning strategy to address both challenges. Our method incorporates progressively updated positive and negative correlation matrices to capture co-occurrence and disjoint relationships within the label space of annotated samples, enabling a holistic assessment of uncertainty rather than treating labels as isolated elements. Furthermore, alongside diversity, our model employs ensemble pseudo labeling and beta scoring rules to address data imbalances. Extensive experiments on four realistic datasets demonstrate that our strategy consistently achieves more reliable and superior performance, compared to several established methods.

Figures

Figures reproduced from arXiv: 2411.17941 by the authors.

Figure 1
Figure 1. Overview of the CRAB framework. It is trained using an ensemble method that generates pseudo labels based [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Averaged micro-F1 score on TextCNN, averaged [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 5
Figure 5. CRAB maintains superior performance across syn [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figures from the paper (5 more)
Figure 7
Figure 7. Figure 7: Trend of hard-to-learn and negative-correlated [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: (a) Performance for different size of hard-to-learn [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 5
Figure 5. Figure 5: Averaged micro-F1 score on synthetic dataset [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: The averaged MeanIR of selected samples, aver [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 9
Figure 9. Figure 9: (a) Ablation study of the asymmetric-correlated la [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 52 canonical work pages

  1. [1]

    A review of uncertainty quantification in deep learning: Techniques, applications and challenges

    Moloud Abdar, Farhad Pourpanah, Sadiq Hussain, Dana Rezazadegan, Li Liu, Mohammad Ghavamzadeh, Paul Fieguth, Xiaochun Cao, Abbas Khosravi, and U Rajendra Acharya. A review of uncertainty quantification in deep learning: Techniques, applications and challenges. Information fusion, 76: 0 243--297, 2021. ISSN 1566-2535

    work page 2021

  2. [2]

    Rebalancing label distribution while eliminating inherent waiting time in multi label active learning applied to transformers

    Maxime Arens, Lucile Callebert, Mohand Boughanem, and Jos \'e G Moreno. Rebalancing label distribution while eliminating inherent waiting time in multi label active learning applied to transformers. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 13621--13632, 2024

    work page 2024

  3. [3]

    Toward label-efficient neural network training: Diversity-based sampling in semi-supervised active learning

    Felix Buchert, Nassir Navab, and Seong Tae Kim. Toward label-efficient neural network training: Diversity-based sampling in semi-supervised active learning. IEEE Access, 11: 0 5193--5205, 2023

    work page 2023

  4. [4]

    Loss functions for binary class probability estimation and classification: Structure and applications

    Andreas Buja, Werner Stuetzle, and Yi Shen. Loss functions for binary class probability estimation and classification: Structure and applications. Working draft, November, 3: 0 13, 2005

    2005

  5. [5]

    Improved multi-label classification under temporal concept drift: Rethinking group-robust algorithms in a label-wise setting

    Ilias Chalkidis and Anders S gaard. Improved multi-label classification under temporal concept drift: Rethinking group-robust algorithms in a label-wise setting. In Findings of the Association for Computational Linguistics: ACL 2022, pages 2441--2454, 2022

    work page 2022

  6. [6]

    Multieurlex-a multi-lingual and multi-label legal document classification dataset for zero-shot cross-lingual transfer

    Ilias Chalkidis, Manos Fergadiotis, and Ion Androutsopoulos. Multieurlex-a multi-lingual and multi-label legal document classification dataset for zero-shot cross-lingual transfer. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 6974--6996, 2021

    work page 2021

  7. [7]

    Active bias: Training more accurate neural networks by emphasizing high variance samples

    Haw-Shiuan Chang, Erik Learned-Miller, and Andrew McCallum. Active bias: Training more accurate neural networks by emphasizing high variance samples. Advances in Neural Information Processing Systems, 30, 2017

    work page 2017

  8. [8]

    Addressing imbalance in multilabel classification: Measures and random resampling algorithms

    Francisco Charte, Antonio J Rivera, Mar \' a J del Jesus, and Francisco Herrera. Addressing imbalance in multilabel classification: Measures and random resampling algorithms. Neurocomputing, 163: 0 3--16, 2015

    work page 2015

Show all 60 references
  1. [9]

    Stable matching-based two-way selection in multi-label active learning with imbalanced data

    Shuyue Chen, Ran Wang, Jian Lu, and Xizhao Wang. Stable matching-based two-way selection in multi-label active learning with imbalanced data. Information Sciences, 610: 0 281--299, 2022

    2022

  2. [10]

    Active learning for bert: an empirical study

    Liat Ein Dor, Alon Halfon, Ariel Gera, Eyal Shnarch, Lena Dankin, Leshem Choshen, Marina Danilevsky, Ranit Aharonov, Yoav Katz, and Noam Slonim. Active learning for bert: an empirical study. In Proceedings of the 2020 conference on empirical methods in natural language process...

    2020

  3. [11]

    Gradient descent finds global minima of deep neural networks

    Simon Du, Jason Lee, Haochuan Li, Liwei Wang, and Xiyu Zhai. Gradient descent finds global minima of deep neural networks. In International conference on machine learning, pages 1675--1685. PMLR, 2019

    2019

  4. [12]

    The lottery ticket hypothesis: Finding sparse, trainable neural networks

    Jonathan Frankle and Michael Carbin. The lottery ticket hypothesis: Finding sparse, trainable neural networks. In International Conference on Learning Representations, 2018

    2018

  5. [13]

    Strictly proper scoring rules, prediction, and estimation

    Tilmann Gneiting and Adrian E Raftery. Strictly proper scoring rules, prediction, and estimation. Journal of the American statistical Association, 102 0 (477): 0 359--378, 2007

    2007

  6. [14]

    An online active multi-label classification algorithm based on a hybrid label query strategy

    Kailun Gong and Tingting Zhai. An online active multi-label classification algorithm based on a hybrid label query strategy. In 2021 3rd International Conference on Machine Learning, Big Data and Business Intelligence (MLBDBI), pages 463--468. IEEE, 2021

    2021

  7. [15]

    Plvi-ce: a multi-label active learning algorithm with simultaneously considering uncertainty and diversity

    Yan Gu, Jicong Duan, Hualong Yu, Xibei Yang, and Shang Gao. Plvi-ce: a multi-label active learning algorithm with simultaneously considering uncertainty and diversity. Applied Intelligence, 53 0 (22): 0 27844--27864, 2023

    2023

  8. [16]

    Feature mixing-based active learning for multi-label text classification

    Xue Han, Qing Wang, Yitong Wang, Jiahui Wang, Chao Deng, and Junlan Feng. Feature mixing-based active learning for multi-label text classification. In ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 10551--10555. IEEE, 2024

    2024

  9. [17]

    Deal: Deep evidential active learning for image classification

    Patrick Hemmer, Niklas Kühl, and Jakob Schöffer. Deal: Deep evidential active learning for image classification. Deep Learning Applications, Volume 3, pages 171--192, 2022. ISSN 9811633568

    2022

  10. [18]

    Uncertainty-based active learning by bayesian u-net for multi-label cone-beam ct segmentation

    Jiayu Huang, Nazbanoo Farpour, Bingjian J Yang, Muralidhar Mupparapu, Fleming Lure, Jing Li, Hao Yan, and Frank C Setzer. Uncertainty-based active learning by bayesian u-net for multi-label cone-beam ct segmentation. Journal of Endodontics, 50 0 (2): 0 220--228, 2024. ISSN 0099-2399

    2024

  11. [19]

    Multi-label classification by exploiting local positive and negative pairwise label correlation

    Jun Huang, Guorong Li, Shuhui Wang, Zhe Xue, and Qingming Huang. Multi-label classification by exploiting local positive and negative pairwise label correlation. Neurocomputing, 257: 0 164--174, 2017

    2017

  12. [20]

    Local positive and negative label correlation analysis with label awareness for multi-label classification

    Rui Huang and Liuyue Kang. Local positive and negative label correlation analysis with label awareness for multi-label classification. International Journal of Machine Learning and Cybernetics, 12 0 (9): 0 2659--2672, 2021

    2021

  13. [21]

    Active query driven by uncertainty and diversity for incremental multi-label learning

    Sheng-Jun Huang and Zhi-Hua Zhou. Active query driven by uncertainty and diversity for incremental multi-label learning. In 2013 IEEE 13th international conference on data mining, pages 1079--1084. IEEE, 2013

    2013

  14. [22]

    Mimic-iii, a freely accessible critical care database

    Alistair EW Johnson, Tom J Pollard, Lu Shen, Li-wei H Lehman, Mengling Feng, Mohammad Ghassemi, Benjamin Moody, Peter Szolovits, Leo Anthony Celi, and Roger G Mark. Mimic-iii, a freely accessible critical care database. Scientific data, 3 0 (1): 0 1--9, 2016

    2016

  15. [23]

    Active learning with complementary sampling for instructing class-biased multi-label text emotion classification

    Xin Kang, Xuefeng Shi, Yunong Wu, and Fuji Ren. Active learning with complementary sampling for instructing class-biased multi-label text emotion classification. IEEE Transactions on Affective Computing, 14 0 (1): 0 523--536, 2020

    2020

  16. [24]

    An exploration of encoder-decoder approaches to multi-label classification for legal and biomedical text

    Yova Kementchedjhieva and Ilias Chalkidis. An exploration of encoder-decoder approaches to multi-label classification for legal and biomedical text. In Anna Rogers, Jordan Boyd-Graber, and Naoaki Okazaki, editors, Findings of the Association for Computational Linguistics: ACL ...

    2023 doi

  17. [25]

    Re-thinking federated active learning based on inter-class diversity

    SangMook Kim, Sangmin Bae, Hwanjun Song, and Se-Young Yun. Re-thinking federated active learning based on inter-class diversity. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3944--3953, 2023

    2023

  18. [26]

    Rcv1: A new benchmark collection for text categorization research

    David D Lewis, Yiming Yang, Tony Russell-Rose, and Fan Li. Rcv1: A new benchmark collection for text categorization research. Journal of machine learning research, 5 0 (Apr): 0 361--397, 2004

    2004

  19. [27]

    Active learning with multi-label svm classification

    Xin Li and Yuhong Guo. Active learning with multi-label svm classification. In IjCAI, volume 13, pages 1479--1485. Citeseer, 2013

    2013

  20. [28]

    Neuralclassifier: an open-source neural hierarchical multi-label text classification toolkit

    Liqun Liu, Funan Mu, Pengyu Li, Xin Mu, Jing Tang, Xingsheng Ai, Ran Fu, Lifeng Wang, and Xing Zhou. Neuralclassifier: an open-source neural hierarchical multi-label text classification toolkit. In Proceedings of the 57th Annual Meeting of the Association for Computational Lin...

    2019

  21. [29]

    Recurrent neural network for text classification with multi-task learning

    Pengfei Liu, Xipeng Qiu, and Xuanjing Huang. Recurrent neural network for text classification with multi-task learning. In Proceedings of the Twenty-Fifth International Joint Conference on Artificial Intelligence, pages 2873--2879, 2016

    2016

  22. [30]

    Influence selection for active learning

    Zhuoming Liu, Hao Ding, Huaping Zhong, Weijia Li, Jifeng Dai, and Conghui He. Influence selection for active learning. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 9274--9283, 2021

    2021

  23. [31]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. In International Conference on Learning Representations, 2019

    2019

  24. [32]

    Multi-label few/zero-shot learning with knowledge aggregated from multiple label graphs

    Jueqing Lu, Lan Du, Ming Liu, and Joanna Dipnall. Multi-label few/zero-shot learning with knowledge aggregated from multiple label graphs. In Empirical Methods in Natural Language Processing 2020, pages 2935--2943. Association for Computational Linguistics (ACL), 2020

    2020

  25. [33]

    Combining graph transformers based multi-label active learning and informative data augmentation for chest xray classification

    Dwarikanath Mahapatra, Behzad Bozorgtabar, Zongyuan Ge, Mauricio Reyes, and Jean-Philippe Thiran. Combining graph transformers based multi-label active learning and informative data augmentation for chest xray classification. In Proceedings of the AAAI Conference on Artificial...

    2024

  26. [34]

    Multi-label active learning through serial--parallel neural networks

    Xue-Yang Min, Kun Qian, Ben-Wen Zhang, Guojie Song, and Fan Min. Multi-label active learning through serial--parallel neural networks. Knowledge-Based Systems, 251: 0 109226, 2022

    2022

  27. [35]

    o llenbrok and Beg \

    Lars M \"o llenbrok and Beg \"u m Demir. Active learning guided fine-tuning for enhancing self-supervised based multi-label classification of remote sensing images. In IGARSS 2023-2023 IEEE International Geoscience and Remote Sensing Symposium, pages 4986--4989. IEEE, 2023

    2023

  28. [36]

    o llenbrok, Gencer Sumbul, and Beg \

    Lars M \"o llenbrok, Gencer Sumbul, and Beg \"u m Demir. Deep active learning for multi-label classification of remote sensing images. IEEE Geoscience and Remote Sensing Letters, 2023

    2023

  29. [37]

    Active learning for hierarchical multi-label classification

    Felipe Kenji Nakano, Ricardo Cerri, and Celine Vens. Active learning for hierarchical multi-label classification. Data Mining and Knowledge Discovery, 34 0 (5): 0 1496--1530, 2020

    2020

  30. [38]

    Pytorch: An imperative style, high-performance deep learning library

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. Pytorch: An imperative style, high-performance deep learning library. Advances in neural information processing systems, 32, 2019

    2019

  31. [39]

    Negative correlation learning in the extreme learning machine framework

    Carlos Perales-Gonz \'a lez, Mariano Carbonero-Ruz, Javier Perez-Rodriguez, David Becerra-Alonso, and Francisco Fern \'a ndez-Navarro. Negative correlation learning in the extreme learning machine framework. Neural Computing and Applications, 32: 0 13805--13823, 2020

    2020

  32. [40]

    Sampling bias in deep active classification: An empirical study

    Ameya Prabhu, Charles Dognin, and Maneesh Singh. Sampling bias in deep active classification: An empirical study. In Kentaro Inui, Jing Jiang, Vincent Ng, and Xiaojun Wan, editors, Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9...

    2019 doi

  33. [41]

    A survey of deep active learning

    Pengzhen Ren, Yun Xiao, Xiaojun Chang, Po-Yao Huang, Zhihui Li, Brij B Gupta, Xiaojiang Chen, and Xin Wang. A survey of deep active learning. ACM computing surveys (CSUR), 54 0 (9): 0 1--40, 2021. ISSN 0360-0300

    2021

  34. [42]

    Toward optimal active learning through sampling estimation of error reduction

    Nicholas Roy and Andrew McCallum. Toward optimal active learning through sampling estimation of error reduction. In ICML, volume 1, page 5. Citeseer, 2001

    2001

  35. [43]

    Distilbert, a distilled version of bert: smaller, faster, cheaper and lighter

    V Sanh. Distilbert, a distilled version of bert: smaller, faster, cheaper and lighter. In Proceedings of Thirty-third Conference on Neural Information Processing Systems (NIPS2019), 2019

    2019

  36. [44]

    A gaussian process-bayesian bernoulli mixture model for multi-label active learning

    Weishi Shi, Dayou Yu, and Qi Yu. A gaussian process-bayesian bernoulli mixture model for multi-label active learning. Advances in Neural Information Processing Systems, 34: 0 27542--27554, 2021

    2021

  37. [45]

    Study of uncertainty quantification using multi-label ecg in deep learning models

    Raquel Simao, Mar \' lia Barandas, David Belo, and Hugo Gamboa. Study of uncertainty quantification using multi-label ecg in deep learning models. In BIOSIGNALS, pages 252--259, 2023

    2023

  38. [46]

    Rethinking deep active learning: Using unlabeled data at model training

    Oriane Siméoni, Mateusz Budnik, Yannis Avrithis, and Guillaume Gravier. Rethinking deep active learning: Using unlabeled data at model training. In 2020 25th International conference on pattern recognition (ICPR), pages 1220--1227. IEEE, 2020. ISBN 1728188083

    2020

  39. [47]

    Cost-efficient multi-instance multi-label active learning via correlation of features

    Guoliang Su, Zhangquan Wu, Yujia Ye, Maoxing Chen, and Jun Zhou. Cost-efficient multi-instance multi-label active learning via correlation of features. In 2023 IEEE International Conference on Image Processing (ICIP), pages 410--414. IEEE, 2023

    2023

  40. [48]

    Harnessing the power of beta scoring in deep active learning for multi-label text classification

    Wei Tan, Ngoc Dang Nguyen, Lan Du, and Wray Buntine. Harnessing the power of beta scoring in deep active learning for multi-label text classification. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 15240--15248, 2024

    2024

  41. [49]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin. Attention is all you need. In 31st Annual Conference on Neural Information Processing Systems (NIPS), volume 30 of Advances in Neural Information Processing Systems, 2017. URL...

    2017

  42. [50]

    Attribute and label distribution driven multi-label active learning

    Min Wang, Tingting Feng, Zhaohui Shan, and Fan Min. Attribute and label distribution driven multi-label active learning. Applied Intelligence, 52 0 (10): 0 11131--11146, 2022

    2022

  43. [51]

    Towards fewer annotations: Active learning via region impurity and prediction uncertainty for domain adaptive semantic segmentation

    Binhui Xie, Longhui Yuan, Shuang Li, Chi Harold Liu, and Xinjing Cheng. Towards fewer annotations: Active learning via region impurity and prediction uncertainty for domain adaptive semantic segmentation. In Proceedings of the IEEE/CVF conference on computer vision and pattern...

    2022

  44. [52]

    Effective multi-label active learning for text classification

    Bishan Yang, Jian-Tao Sun, Tengjiao Wang, and Zheng Chen. Effective multi-label active learning for text classification. In Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining, pages 917--926, 2009

    2009

  45. [53]

    Not all out-of-distribution data are harmful to open-set active learning

    Yang Yang, Yuxuan Zhang, Xin Song, and Yi Xu. Not all out-of-distribution data are harmful to open-set active learning. Advances in Neural Information Processing Systems, 36, 2024

    2024

  46. [54]

    Rethinking the value of labels for improving class-imbalanced learning

    Yuzhe Yang and Zhi Xu. Rethinking the value of labels for improving class-imbalanced learning. Advances in neural information processing systems, 33: 0 19290--19301, 2020

    2020

  47. [55]

    An overview of overfitting and its solutions

    Xue Ying. An overview of overfitting and its solutions. In Journal of physics: Conference series, volume 1168, page 022022. IOP Publishing, 2019

    2019

  48. [56]

    Cmal: Cost-effective multi-label active learning by querying subexamples

    Guoxian Yu, Xia Chen, Carlotta Domeniconi, Jun Wang, Zhao Li, Zili Zhang, and Xiangliang Zhang. Cmal: Cost-effective multi-label active learning by querying subexamples. IEEE Transactions on Knowledge and Data Engineering, 34 0 (5): 0 2091--2105, 2020

    2020

  49. [57]

    A sensitivity analysis of (and practitioners’ guide to) convolutional neural networks for sentence classification

    Ye Zhang and Byron C Wallace. A sensitivity analysis of (and practitioners’ guide to) convolutional neural networks for sentence classification. In Proceedings of the Eighth International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 253--263, 2017

    2017

  50. [58]

    Granular multilabel batch active learning with pairwise label correlation

    Yuanjian Zhang, Tianna Zhao, Duoqian Miao, and Witold Pedrycz. Granular multilabel batch active learning with pairwise label correlation. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 52 0 (5): 0 3079--3091, 2021

    2021

  51. [59]

    Uncertainty in bayesian deep label distribution learning

    Rui Zheng, Shulin Zhang, Lei Liu, Yuhao Luo, and Mingzhai Sun. Uncertainty in bayesian deep label distribution learning. Applied Soft Computing, 101: 0 107046, 2021

    2021

  52. [60]

    Addressing the item cold-start problem by attribute-driven active learning

    Yu Zhu, Jinghao Lin, Shibi He, Beidou Wang, Ziyu Guan, Haifeng Liu, and Deng Cai. Addressing the item cold-start problem by attribute-driven active learning. IEEE Transactions on Knowledge and Data Engineering, 32 0 (4): 0 631--644, 2019

    2019

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.