Pith. sign in

REVIEW 4 major objections 5 minor 26 references

One-Class Domain Adaptation via Meta-Learning

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A model can adapt to a new domain using only a handful of normal-class examples, and doing so via meta-learning outperforms standard MAML.

desk verdict Solid empirical paper with a real but modest new setting; the theory oversells cross-domain transfer, but the results warrant a serious referee. read the letter →

arxiv 2501.13052 v1 pith:EYGLPGG3 submitted 2025-01-22 cs.LG

Stephanie Holly , Thomas Bierweiler , Stefan von Dosky , Ahmed Frikha , Clemens Heitzinger , Jana Eder This is my paper · ORCID
classification cs.LG
keywords one-classdomainadaptationmeta-learningMAMLtasksamplingstrategydistributionshiftanomalyclassificationfew-shotlearningindustrialIoT
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 formulates the one-class domain adaptation (OC-DA) problem: a multi-class classifier must adapt to a new domain using only a few labeled examples of a single 'normal' class, with no examples of the anomalous or other classes. It proposes a task sampling strategy—support sets contain only normal-class examples while query sets stay class-balanced—that converts any bi-level meta-learning algorithm, in particular MAML, into an OC-DA algorithm. The authors argue, via a Taylor-expansion analysis, that this setup optimizes meta-parameters so that a gradient step on normal-class data improves performance on all classes. On Rainbow-MNIST and on vibration readings from centrifugal pumps, the resulting OC-DA MAML substantially outperforms standard MAML and approaches in-distribution accuracy. The work matters because industrial deployment often has abundant normal data but scarce labeled anomaly data in each new environment.

What carries the argument

The load-bearing mechanism is the one-class task sampling strategy inside a bi-level meta-learning loop. In each meta-training task, the support set $S_i$ contains $K$ examples of only the normal class $n$, while the query set $Q_i$ is class-balanced with $K$ examples per class. The inner adaptation $\mathrm{Alg}(\theta, S_i) = \theta - \alpha \nabla L_{S_i}(\theta)$ is a gradient step on normal-class data, and the outer update differentiates through this step on the query loss. The theoretical analysis uses the second-order Taylor expansion of the MAML gradient (following [18]) to show the meta-objective is equivalent to minimizing the query loss while maximizing the inner product $\langle \nabla L_Q, \nabla L_S \rangle$; this is the precise mechanism that is claimed to make normal-class gradients informative for all other classes within a domain.

What would settle it

Construct a domain shift where the normal-class distribution is identical between source and target but the anomaly-class distributions change arbitrarily; if OC-DA MAML's target accuracy then does not beat a model fine-tuned on the same K normal examples from a random initialization, the claimed transfer mechanism is refuted.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that cross-domain adaptation for a multi-class classifier can be driven entirely by normal-class examples. The proposed OC-DA MAML algorithm samples meta-training tasks with K-shot support sets restricted to the normal class and class-balanced query sets, then runs the standard MAML bi-level loop: an inner gradient step on the normal-only support set produces domain-specific parameters, and the outer update optimizes the meta-initialization so that this step also reduces the loss on the class-balanced query set. The Taylor expansion of the MAML gradient shows the outer objective maximizes the inner product of the support-set and query-set gradients, meaning a gradient step on normal data is trained to improve classification of all other classes. The empirical claim is that after adapting on K normal examples, OC-DA MAML reaches 82.9% (K=1) and 88.4% (K=3) average accuracy on Rainbow-MNIST target domains versus 35.6% and 38.4% for standard MAML, and 92.8% versus 84.3% on the centrifugal-pump dataset, close to the 95.5% in-distribution baseline.

Load-bearing premise

The method works only if a gradient computed on normal-class examples carries information about how to classify the other classes in that domain, which the theoretical analysis assumes and the experiments demonstrate but does not guarantee.

Editorial extensions

If this is right

  • OC-DA MAML reaches 82.9% average accuracy on Rainbow-MNIST target domains after a single normal example per domain, versus 35.6% for standard MAML.
  • On the centrifugal-pump vibration dataset, OC-DA MAML averages 92.8% across the eight source–target splits, versus 84.3% for MAML and 82.2% for the no-adaptation OOD baseline.
  • Target-domain performance after one-class adaptation approaches in-distribution accuracy: 82.9% versus 83.93% on Rainbow-MNIST and 92.8% versus 95.5% on the pumps.
  • The task sampling strategy is algorithm-agnostic: it can be applied to any bi-level meta-learning algorithm, not only MAML.
  • Increasing the number of normal-class adaptation examples K from 1 to 3 raises average Rainbow-MNIST accuracy from 82.9% to 88.4%.

Reading between the lines

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

  • The success of the method suggests that in many industrial settings, labeled normal data can serve as a cheap and abundant signal for domain adaptation, reducing the need for costly anomaly data collection in every new environment.
  • The theory's inner-product condition implies a testable prediction: OC-DA gains should be largest when normal-class and anomaly-class features are correlated within a domain, and smallest when they are independent; datasets engineered to violate this could bound the method's applicability.
  • Because the query set is class-balanced during meta-training while deployment anomaly priors are usually imbalanced, the reported accuracies may overstate performance in settings where anomaly classes are rare; evaluating with deployment-like priors is a natural next step.
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 introduces the one-class domain adaptation (OC-DA) problem, in which a multi-class classifier must adapt to a new domain using only a few examples of one designated normal class. It proposes a task-sampling strategy that can be applied to any bi-level meta-learning algorithm, instantiates it as OC-DA MAML, and gives a Taylor-expansion analysis of the resulting meta-update. The method is evaluated on a Rainbow-MNIST benchmark and on an industrial centrifugal-pump vibration dataset, reporting large average accuracy gains over standard MAML on target domains (e.g., 82.9% vs. 35.6% for K=1 on Rainbow-MNIST).

Significance. The problem setting is practically motivated and the reported empirical gains are substantial, especially on the real-world pump dataset where OC-DA MAML reaches accuracy close to in-distribution performance. If the central claim is robust, the paper offers a useful recipe for adapting meta-learning methods to settings where only normal-class data is available at deployment time, which is relevant for industrial IoT anomaly classification. The authors are also explicit about the key empirical premise that normal-class gradients carry domain-specific structure shared with anomalous classes, and Fig. 3 provides visual support for that premise on the pump data. However, the theoretical analysis does not establish cross-domain transfer, and the experimental comparison has fairness and statistical-reporting issues that need to be addressed before the headline claim is fully supported.

major comments (4)
  1. [2.5, Eq. (7)–(8)] The theoretical analysis is the standard second-order Taylor expansion of the MAML gradient and does not establish the OC-DA transfer claim. The derivation shows that, on a given training task, the meta-objective encourages a positive inner product between the normal-only support gradient and the class-balanced query gradient, but it provides no statement about held-out target domains. This is not an OC-DA-specific mechanism, and it relies on the assumption stated in Section 2.2 that meta-training and meta-testing tasks are drawn from the same task distribution, which the OC-DA setting explicitly violates by training on source domains and testing on target domains. The paper should either provide an explicit domain-generalization analysis with conditions under which normal-only adaptation transfers, or substantially soften the claim that the analysis demonstrates cross-domain one-class adaptation.
  2. [Table 5 and Appendix A.2] The comparison between MAML and OC-DA MAML is handicapped by a train/test task mismatch for the MAML baseline. In MAML, the meta-training support set contains K examples from every class, but at meta-testing time the support set contains K examples from the normal class only, as shown in Table 5. Standard MAML was not trained for this support distribution, so part of the reported gap in Tables 2 and 3 may reflect this mismatch rather than the benefit of the OC-DA sampling strategy. A fair comparison would train the MAML baseline with the same normal-only support distribution during meta-training, or report an additional baseline that is matched in this respect.
  3. [Tables 2 and 3] The paper states in Section 3.1 that all results are averaged over three runs with different seeds, but no standard deviations, confidence intervals, or significance tests are reported. Given the empirical claim of 'significantly improves' in the abstract and Section 4, the absence of any statistical measure is load-bearing, especially because the MAML baseline numbers vary considerably across rows (e.g., 74.03 to 92.47 in Table 3). The authors should report per-run variability and, if appropriate, a paired significance test across the evaluated source/target combinations.
  4. [Section 2.5 and Fig. 3] The paper's central premise—that normal-class gradients are informative about all classes in a new domain—is supported only by visual inspection of one real-world dataset and by the within-task gradient-alignment argument, not by a formal guarantee or by broad empirical evidence. The conclusion in Section 4 overstates the generality by claiming that one-class adaptation across domains is enabled. The paper should either provide a formal condition under which the premise holds, or explicitly scope the claim to settings where normal data and anomalous data share domain-specific structure, as observed in the centrifugal-pump data.
minor comments (5)
  1. [Table 4 heading] The heading contains a typo: 'hyerparameters' should be 'hyperparameters'.
  2. [Tables 1 and 3] The text uses 'ODD' where it should be 'OOD' (out-of-distribution); this occurs in Table 1 and in the surrounding discussion in Section 3.2.
  3. [Section 1, contribution (4)] The contribution list refers to 'OC-MAML' whereas the rest of the paper uses 'OC-DA MAML'; please make the notation consistent.
  4. [Section 3.2] The phrase 'significantly outperforming MAML with classical K-shot learning tasks' is ambiguous because the meta-testing support for MAML in this paper is not a classical K-shot support (it contains K normal examples and zero examples of other classes); clarify what 'classical' means here.
  5. [Appendix A.1] The text says the Centrifugal-Pumps data provides 32 domains and then reports 12/4/4 splits, which totals 20 domains; the relationship between the 32 domains and the 12/4/4 split should be explained more clearly.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central OC-DA MAML claims are tested on held-out target domains, and the theory section is a standard MAML gradient expansion rather than a derivation that secretly assumes its conclusion.

full rationale

The paper's central claims are empirical: OC-DA MAML outperforms standard MAML on held-out target domains (Table 2 for Rainbow-MNIST, Table 3 for Centrifugal-Pumps). These results are obtained by adapting on K normal-class examples in target domains that were not used for meta-training; they are not fitted parameters renamed as predictions, and evaluation uses class-balanced query sets separate from the adaptation supports. The theoretical analysis in Sec. 2.5 reproduces the standard second-order Taylor expansion of the MAML gradient (Eqs. 7-8), citing Nichol et al. [18]. It shows that the meta-update minimizes query loss and maximizes the inner product between support and query gradients. Since OC-DA MAML's support is normal-only and its query is class-balanced, the analysis is an interpretation of why the objective could encourage one-class-to-all-class generalization within a domain. The paper then informally extends this to 'across domains,' but that step is an unsupported generalization about transfer, not a circular reduction: no equation in Sec. 2.5 uses target-domain data or assumes the conclusion. The self-citations to FSOC [5] and the pump dataset [25] are related-work and data-provenance references; FSOC is an independently published AAAI paper and the sampling strategy is explicitly adapted rather than disguised. Accordingly, there is no load-bearing self-citation or definitional equivalence that makes the results forced by construction. The weakness of Sec. 2.5 as a proof of cross-domain transfer is a rigor concern, not a circularity concern.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central method introduces no fitted constants in the derivation itself; the free parameters are standard training hyperparameters chosen by hand. The main load-bearing axioms are the standard meta-learning task-distribution assumption and the empirical premise that normal-class gradients transfer to other classes within a domain.

free parameters (5)
  • Inner learning rate alpha = 0.01
    Hand-chosen for both datasets; no sensitivity analysis is reported, and the Taylor analysis assumes alpha is small.
  • Outer learning rate beta = 0.001
    Hand-chosen; affects convergence of the meta-update.
  • Number of inner gradient steps k = 1
    Fixed to one step; the theoretical derivation considers only one inner step.
  • Meta-batch size |I| = 4 (Rainbow-MNIST), 2 (Centrifugal-Pumps)
    Hand-chosen; no ablation is provided.
  • Shots per class K = 1 or 3 (Rainbow-MNIST), 2 (Centrifugal-Pumps)
    Hand-chosen; results vary with K.
assumptions (4)
  • domain assumption Meta-training and meta-testing tasks are drawn from the same task distribution.
    Invoked in Section 2.2 before Eq. 2; in OC-DA this means the relationship between normal-class support and class-balanced queries transfers across source and target domains.
  • standard math The loss is sufficiently smooth for the second-order Taylor expansion of the MAML gradient to be valid at alpha = 0.01.
    Section 2.5 uses the expansion from [18] with O(alpha^2) error; no verification of smallness or smoothness is provided.
  • ad hoc to paper Normal-class data encodes domain structure shared with anomalous classes.
    Motivated by Fig. 3 for the pump dataset, but the method's transfer guarantee depends on this property holding in unseen target domains.
  • ad hoc to paper Maximizing the inner product between support and query gradients is sufficient for one-class adaptation to work.
    Section 2.5 treats the standard MAML gradient decomposition as the theoretical justification; this is an interpretive application of [18] without a formal guarantee.

how reviews work

0 comments
Cite this review

Pith. "Pith review of One-Class Domain Adaptation via Meta-Learning." pith.science (2026). https://pith.science/paper/EYGLPGG3

@misc{pith2026250113052,
  author       = {Pith},
  title        = {Pith review of: One-Class Domain Adaptation via Meta-Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EYGLPGG3}},
  note         = {Machine review of arXiv:2501.13052}
}
read the original abstract

The deployment of IoT (Internet of Things) sensor-based machine learning models in industrial systems for anomaly classification tasks poses significant challenges due to distribution shifts, as the training data acquired in controlled laboratory settings may significantly differ from real-time data in production environments. Furthermore, many real-world applications cannot provide a substantial number of labeled examples for each anomalous class in every new environment. It is therefore crucial to develop adaptable machine learning models that can be effectively transferred from one environment to another, enabling rapid adaptation using normal operational data. We extended this problem setting to an arbitrary classification task and formulated the one-class domain adaptation (OC-DA) problem setting. We took a meta-learning approach to tackle the challenge of OC-DA, and proposed a task sampling strategy to adapt any bi-level meta-learning algorithm to OC-DA. We modified the well-established model-agnostic meta-learning (MAML) algorithm and introduced the OC-DA MAML algorithm. We provided a theoretical analysis showing that OC-DA MAML optimizes for meta-parameters that enable rapid one-class adaptation across domains. The OC-DA MAML algorithm is evaluated on the Rainbow-MNIST meta-learning benchmark and on a real-world dataset of vibration-based sensor readings. The results show that OC-DA MAML significantly improves the performance on the target domains and outperforms MAML using the standard task sampling strategy.

Figures

Figures reproduced from arXiv: 2501.13052 by the authors.

Figure 1
Figure 1. Example of a 3-way 1-shot learning task in the OC-DA setting [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. illustrates differences in the data distribution across two different domains in the Centrifugal-Pumps dataset. The plots show the average amplitude of FFT vibration signals for each class (1-normal, 5-cavitation, 6-hydraulic blockage, 7- dry running). The data was recorded by the same pump placed on two different surfaces: within a steel framework and on a concrete surface. We observe sig￾nificant differences in th… view at source ↗
Figure 3
Figure 3. Visualization of domain-specific information in the Centrifugal-Pumps dataset, showing the average amplitude [mm] of the FFT vibration signals per frequency [Hz] with a 95% confidence interval. Each anomalous class (5-cavitation, 6-hydraulic block￾age, 7-dry running) is compared to the normal class (1-normal class), recorded by same pump operated within a steel framework vs. on a concrete surface [PITH_FULL_IMAGE:f… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 22 canonical work pages

  1. [1]

    Invariant Risk Minimization

    Arjovsky Martin, Bottou Léon, Gulrajani Ishaan, and Lopez-Paz David. Invariant Risk Minimization. InarXiv preprint arXiv:1907.02893, 2019

    arXiv 1907

  2. [2]

    Learn2learn: A Library for Meta-Learning Research

    Arnold Sébastien M.R., Mahajan Praateek, Datta Debajyoti, Bunner Ian, and Zarkias Konstantinos Saitas. Learn2learn: A Library for Meta-Learning Research. In arXiv preprint arXiv:2008.12284, 2020

    arXiv 2008

  3. [3]

    Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks

    Finn Chelsea, Abbeel Pieter, and Levine Sergey. Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks. InThe Proceedings of the 34th International Conference on Machine Learning (ICML), volume 70, pages 1126–1135, Sydney, Australia, 2017. PMLR

    work page 2017

  4. [4]

    Online Meta-Learning

    Finn Chelsea, Rajeswaran Aravind, Kakade Sham, and Levine Sergey. Online Meta-Learning. In The Proceedings of the 36th International Conference on Ma- chine Learning (ICML), volume 97, pages 1920–1930, Long Beach, USA, 2019. PMLR

    work page 1920

  5. [5]

    Few-Shot One-Class Classification via Meta-Learning

    Frikha Ahmed, Krompaß Denis, Köpken Hans-Georg, and Tresp Volker. Few-Shot One-Class Classification via Meta-Learning. InThe Proceedings of the 35th AAAI Conference on Artificial Intelligence, 35(8):7448–7456, 2021. 14 S. Holly et al

    work page 2021

  6. [6]

    ARCADe: A Rapid Continual Anomaly Detector

    Frikha Ahmed, Krompaß Denis, Tresp Volker. ARCADe: A Rapid Continual Anomaly Detector. 2020 25th International Conference on Pattern Recognition (ICPR), pages 10449–10456, 2020

    work page 2020

  7. [7]

    Discovery of New Multi-Level Features for Domain Generalization via Knowledge Corruption

    FrikhaAhmed,KrompaßDenis,andTrespVolker.Columbus:Automateddiscovery of new multi-level features for domain generalization via knowledge corruption. In arXiv preprint arXiv:2109.04320, 2021

    work page Pith review arXiv 2021

  8. [8]

    TransferLearningStrategiesforAnomalyDetectioninIoTVibrationData

    Heistracher Clemens, Jalali Anahid, Strobl Indu, Suendermann Axel, Meixner Se- bastian, Holly Stephanie, Schall Daniel, Haslhofer Bernhard, and Kemnitz Jana. TransferLearningStrategiesforAnomalyDetectioninIoTVibrationData. IECON 2021 – 47th Annual Conference of the IEEE Industrial Electronics Society, pages 1–6, Toronto, Canada, 2021

    work page 2021

Show all 26 references
  1. [9]

    Meta- Learning in Neural Networks: A Survey.IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(9), 5149–5169, 2022

    Hospedales Timothy, Antoniou Antreas, Micaelli Paul, and Storkey Amos. Meta- Learning in Neural Networks: A Survey.IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(9), 5149–5169, 2022

    2022

  2. [10]

    Universal Language Model Fine-tuning for Text Classification

    Howard Jeremy and Ruder Sebastian. Universal Language Model Fine-tuning for Text Classification. InarXiv preprint arXiv:1801.06146, 2018

    2018 arXiv

  3. [11]

    Towards Robust and Transferable IIoT Sensor based Anomaly Classifica- tion using Artificial Intelligence

    Kemnitz Jana, Bierweiler Thomas, Grieb Herbert, van Dosky Stefan, and Schall Daniel. Towards Robust and Transferable IIoT Sensor based Anomaly Classifica- tion using Artificial Intelligence. InProceedings of the 4th International Data Sci- ence Conference – iDSC2021, Data Scie...

    2022

  4. [12]

    An Edge Deployment Framework to Scale AI in Industrial Applications.2023 IEEE 7th International Conference on Fog and Edge Computing (ICFEC), pages 24–32, Bangalore, India, 2023

    Kemnitz Jana, Weissenfeld Axel, Schoeffl Leopold, Stiftinger Andreas, Rechberger Daniel, Prangl Bernhard, Kaufmann Thomas, Hiessl Thomas, Holly Stephanie, Heistracher Clemens, and Schall Daniel. An Edge Deployment Framework to Scale AI in Industrial Applications.2023 IEEE 7th ...

    2023

  5. [13]

    Siamese Neural Net- works for One-Shot Image Recognition.Deep Learning Workshop at the 32nd In- ternational Conference on Machine Learning (ICML), 2(1):1–30, 2015

    Koch Gregory, Zemel Richard, and Salakhutdinov Ruslan. Siamese Neural Net- works for One-Shot Image Recognition.Deep Learning Workshop at the 32nd In- ternational Conference on Machine Learning (ICML), 2(1):1–30, 2015

    2015

  6. [14]

    WILDS: A Benchmark of in-the-Wild Distribution Shifts

    Koh Pang Wei et al. WILDS: A Benchmark of in-the-Wild Distribution Shifts. In The Proceedings of the 38th International Conference on Machine Learning (ICML), volume 139, pages 5637–5664, 2021. PMLR

    2021

  7. [15]

    Identification of Inlet Pipe Blockage Level in Centrifugal Pump over a Range of Speeds by Deep Learning Algorithm using Multi-Source Data.Measurement, 186(110146), 2021

    Kumar Dhiraj, Dewangan Aakash, Tiwari Rajiv, and Bordoloi D.J. Identification of Inlet Pipe Blockage Level in Centrifugal Pump over a Range of Speeds by Deep Learning Algorithm using Multi-Source Data.Measurement, 186(110146), 2021

    2021

  8. [16]

    Online Meta-Learning for Multi-Source and Semi- Supervised Domain Adaptation

    Li Da and Hospedales Timothy. Online Meta-Learning for Multi-Source and Semi- Supervised Domain Adaptation. InComputer Vision – ECCV 2020, Lecture Notes in Computer Science, volume 12361, pages 382–403. Springer, 2020

    2020

  9. [17]

    Learning to Gener- alize: Meta-Learning for Domain Generalization

    Li Da, Yang Yongxin, Song Yi-Zhe and Hospedales Timothy. Learning to Gener- alize: Meta-Learning for Domain Generalization. InThe Proceedings of the 32nd AAAI Conference on Artificial Intelligence, 32(1):3490–3497, 2018

    2018

  10. [18]

    On First-Order Meta-Learning Algorithms

    Nichol Alex, Achiam Joshua, and Schulman John. On First-Order Meta-Learning Algorithms. InarXiv preprint arXiv:1803.02999, 2018

    2018 arXiv

  11. [19]

    Meta- Learning with Implicit Gradients

    Rajeswaran Aravind, Finn Chelsea, Kakade Sham, and Levine Sergey. Meta- Learning with Implicit Gradients. In The Proceedings of the 33rd International Conference on Neural Information Processing Systems (NeurIPS), 32(11):113–124, 2019

    2019

  12. [20]

    Optimization as a Model for Few-Shot Learning

    Ravi Sachin and Larochelle Hugo. Optimization as a Model for Few-Shot Learning. In The 5th International Conference on Learning Representations (ICLR) Confer- ence Track Proceedings, 2017. One-Class Domain Adaptation via Meta-Learning 15

    2017

  13. [21]

    Imagenet Large Scale Visual Recognition Challenge.In- ternational Journal of Computer Vision, volume 115, pages 211–252, 2015

    Russakovsky Olga et al. Imagenet Large Scale Visual Recognition Challenge.In- ternational Journal of Computer Vision, volume 115, pages 211–252, 2015

    2015

  14. [22]

    Prototypical Networks for Few- Shot Learning

    Snell Jake, Swersky Kevin, and Zemel Richard. Prototypical Networks for Few- Shot Learning. InThe Proceedings of the 31st International Conference on Neural Information Processing Systems (NeurIPS), volume 30, pages 4080–4090, 2017

    2017

  15. [23]

    Vinyals Oriol, Blundell Charles, Lillicrap Timothy, Kavukcuoglu Koray, and Wier- stra Daan. Matching Networks for One Shot Learning, In The Proceedings of the 30th International Conference on Neural Information Processing Systems (NeurIPS), volume 29, pages 3637–3645, 2016

    2016

  16. [24]

    In The 10th International Conference on Learning Representations (ICLR), 2022

    YaoHuaxiu,andZhangLinjun,andFinnChelsea.Meta-LearningwithFewerTasks through Task Interpolation. In The 10th International Conference on Learning Representations (ICLR), 2022

    2022

  17. [25]

    Autoencoder Based Anomaly Detection and Ex- plained Fault Localization in Industrial Cooling Systems

    Holly Stephanie, Heel Robin, Katic Denis, Schoeffl Leopold, Stiftinger Andreas, Holzner Peter, Kaufmann Thomas, Haslhofer Bernhard, Schall Daniel, Heitzinger Clemens, and Kemnitz Jana. Autoencoder Based Anomaly Detection and Ex- plained Fault Localization in Industrial Cooling...

    2022

  18. [26]

    Evaluation of Hyperparameter- Optimization Approaches in an Industrial Federated Learning System

    Holly Stephanie, Hiessl Thomas, Lakani Safoura Rezapoura, Schall Daniel, Heitzinger Clemens, and Kemnitz Jana. Evaluation of Hyperparameter- Optimization Approaches in an Industrial Federated Learning System. InProceed- ings of the 4th International Data Science Conference – i...

    2022

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.