Pith. sign in

REVIEW 5 major objections 5 minor 1 cited by

Evaluating deep learning models for fault diagnosis of a rotating machinery with epistemic and aleatoric uncertainty

T0 review · 5 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Deep ensemble architectures—especially a heterogeneous ensemble called De2—are the most reliable uncertainty-aware models for fault diagnosis in rotating machinery, detecting out-of-distribution data better and faster than Monte Carlo…

desk verdict A useful benchmark with a practical new threshold, but the ensemble advantage is confounded by base architecture and needs a deterministic CNN control. read the letter →

arxiv 2412.18980 v2 pith:TEGGJ5ZC submitted 2024-12-25 cs.LG

Reza Jalayer , Masoud Jalayer , Andrea Mor , Carlotta Orsenigo , Carlo Vercellis This is my paper · ORCID
classification cs.LG
keywords FaultDiagnosisUncertaintyTrustworthyAIIndustry4.0DeepLearningOutofDistributionDataensemblesepistemicandaleatoric
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 compares four uncertainty-aware deep learning architectures on a bearing fault diagnosis benchmark to see which can best flag data that falls outside the training distribution—either because it comes from an unseen fault (epistemic uncertainty) or because sensor noise corrupts the signal (aleatoric uncertainty). The authors find that deep ensembles, and particularly a heterogeneous ensemble they call De2, detect out-of-distribution data more reliably than Monte Carlo dropout or a Bayesian neural network under both kinds of uncertainty, while also making faster predictions. They also propose a new entropy threshold, based on F1 score, that reduces the number of trustworthy samples wrongly flagged as untrustworthy compared with the standard outlier-based threshold. If the finding holds, deep ensembles become the practical default for trustworthy fault diagnosis in industrial settings.

What carries the argument

The central machinery is the prediction-entropy test: each model produces K predictions per example (K=10 for dropout and BNN, K=4 for ensembles), the average softmax scores are converted to entropy, and an entropy threshold decides whether the example is trustworthy (in-distribution) or untrustworthy (out-of-distribution). The paper compares two thresholds: τ1, the inter-quartile-range outlier rule applied to in-distribution validation entropies, and τ2, a new threshold chosen to maximize F1 score on the validation confusion matrix. The architecture that carries the argument is De2, a four-learner deep ensemble with two simple CNN learners and two hybrid ConvLSTM learners, whose diversity is credited with the most accurate uncertainty quantification.

What would settle it

Train a deterministic CNN with the same capacity as a single De2 learner on CWRU and measure its OOD detection under the same thresholds; if it reaches De2's roughly 90–97% accuracy, the ensemble benefit is an artifact of backbone choice rather than the ensembling mechanism. Alternatively, average the predictions of three independently seeded ConvLSTM-D models and compare with De2.

Watch

Extended reading notes

Core claim

On the CWRU rotating-bearing benchmark, across six epistemic scenarios (each holding out one fault class) and twelve aleatoric scenarios (four noise types at three signal-to-noise ratios), the deep ensemble models De1 and De2 outperform ConvLSTM with MC dropout and a Bayesian neural network at separating in-distribution from out-of-distribution data by prediction entropy. De2, combining two simple CNNs with two ConvLSTM-based learners, achieves the top average OOD detection rates—97.5% and 89.1% under the two thresholds for epistemic uncertainty, and up to 100% under strong noise—while keeping in-distribution misclassification low and delivering the shortest prediction time. The paper also introduces an F1-score-based entropy threshold (τ2) that is consistently higher than the existing IQR-based threshold (τ1), trading a small loss in OOD recall for a substantial reduction in false alarms on in-distribution data.

Load-bearing premise

The comparison assumes each method's performance reflects its uncertainty mechanism, because the dropout, Bayesian, and ensemble models are built on different neural network backbones.

Editorial extensions

If this is right

  • If De2's advantage generalizes, practitioners deploying uncertainty-aware fault diagnosis should prefer heterogeneous deep ensembles over dropout sampling or BNNs, especially when unseen fault classes are likely.
  • The choice of entropy threshold matters: conservative τ1 maximizes OOD recall at the cost of more false alarms, while τ2 minimizes manual re-inspection burden because it flags fewer in-distribution samples.
  • Under slight noise, all methods struggle to distinguish noisy OOD from clean ID data, so noise should be controlled at data-collection time rather than only at inference.
  • Deep ensembles' faster prediction time means earlier fault detection and intervention, despite longer training.
  • Non-Gaussian noise types (Weibull, Impulse, Rayleigh) matter: detection performance differs by noise type, so evaluations restricted to Gaussian noise may misstate real-world performance.

Reading between the lines

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

  • The paper's recommendation presupposes that the observed advantage comes from ensembling itself, since each method uses a different backbone; a natural test is to compare a single strong CNN with a deep ensemble built on the same backbone.
  • The F1-based threshold τ2 could be transferred to other OOD detection settings beyond fault diagnosis, wherever a validation set contains both ID and OOD samples; its behavior under class imbalance is worth probing.
  • Since the paper assumes the noise type is known at deployment, a practical extension would be a noise-type classifier upstream, or training with mixed noise types, to remove that assumption.
  • The 20% noise-injection protocol means the detector never sees pure-signal baselines for those OOD samples; real signals might mix clean and noisy segments within a single burst, which could change optimal thresholds.
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

5 major / 5 minor

Summary. The paper reports an empirical comparison of four uncertainty-aware deep learning architectures for bearing fault diagnosis on the CWRU dataset: Monte Carlo dropout on a ConvLSTM (ConvLSTM-D), a Bayesian neural network (BNN), and two deep ensembles (De1 and De2). Six epistemic scenarios are constructed by holding out one fault class, and twelve aleatoric scenarios by adding Gaussian, Impulse, Rayleigh, and Weibull noise at three SNR levels. OOD detection is based on entropy thresholds: the IQR-based tau1 and a proposed F1-maximizing threshold tau2. The main finding is that De2 has the highest OOD detection rates and, under tau2, the lowest ID misclassification, especially under strong noise, and the shortest prediction time; the authors conclude that deep ensembles, particularly De2, are the preferred choice for uncertainty-aware fault diagnosis.

Significance. If the empirical ranking is trustworthy, the paper provides useful practical guidance and broadens the aleatoric-uncertainty evaluation to non-Gaussian noise types. The proposed tau2 threshold is a simple and practical alternative to tau1, and the central ranking is evaluated on a separate test set, which avoids the circularity failure mode of fitting a threshold and reporting performance on the same data. The paper's strength is its breadth of scenarios and systematic reporting of OOD/ID rates; its weakness is the lack of experimental controls needed for a comparative benchmark. The manuscript does not ship machine-checked proofs or code, and the main recommendation rests on comparisons in which the uncertainty mechanism and the base architecture vary together.

major comments (5)
  1. [Section 3.2, Tables 5-7] The comparison conflates the uncertainty mechanism with the base architecture. MC dropout is implemented on a ConvLSTM (ConvLSTM-D), BNN on a convolutional stack with Bayesian dense layers, De1 on four identical small CNNs, and De2 on a mixture of CNN and ConvLSTM learners; no deterministic single-model baseline shares the architecture of a De1 or De2 base learner. The higher OOD detection of De1/De2 (e.g., 93.9% and 97.5% vs. 71.4% and 70.4% under tau1 in Table 6) could therefore be due to architecture rather than ensembling. The very low tau1 values for De1/De2 in Table 5 (0.0001 in several scenarios) reinforce this risk, because a model that is extremely confident on ID data will flag almost any OOD point under an IQR-based threshold. Please add at least one deterministic CNN baseline with the same architecture as the De1 base learner, and ideally a single ConvLSTM baseline matching De2's LSTM-based learners, so that the ensemble advantage is measured relative to the same architecture.
  2. [Section 3.1, Figure 7] The data-preparation procedure leaks information between training and test sets. Bursts of 512 points are extracted every 200 points, so consecutive bursts overlap by 312 samples; a random 70/30 split of these bursts places overlapping windows from the same continuous bearing signal in both training and test partitions. This can inflate ID accuracy and artificially separate ID and OOD entropy distributions, and it also affects the validation-based threshold computation. Please use a disjoint split by time intervals or another non-overlapping assignment, and restate the conclusions if the numbers change.
  3. [Sections 4.1-4.2, Tables 6-10] All results are point estimates from a single run; the tables report no confidence intervals, no multiple seeds, and no significance tests. Several claims rely on small margins or scenario-dependent reversals, such as the De1 vs. De2 comparison under tau1 in Tables 6 and 7, so the ranking cannot be assessed statistically as presented. Please add repeated-seed results with confidence intervals or pairwise tests, or explicitly state the single-run limitation and temper the ranking claims.
  4. [Appendix Table A2 vs. Section 3.2.2/Figure 10] The BNN architecture is described inconsistently. The text and figure describe Conv+Pool blocks followed by flattening and Bayesian dense layers with no LSTM, but Table A2 lists an LSTM layer with 64 nodes before the Bayesian dense layer. This ambiguity makes the BNN implementation irreproducible and should be resolved; the same issue affects the description of De2's LSTM-based learners, which the text calls two dense layers but Table A4 lists as a single dense layer after the LSTM.
  5. [Section 4.3, Table 11] The first sentence of Section 4.3 contradicts Table 11. The text says training for De1 and De2 is significantly faster than for ConvLSTM-D and BNN, but Table 11 shows the opposite: De1 and De2 take 688.44 and 639.69 seconds versus 239.55 and 266.18 seconds for ConvLSTM-D and BNN. The following sentence and Section 5 correctly state that prediction is faster for the ensembles while training is longer; please correct this internal contradiction.
minor comments (5)
  1. [Tables 8-10] The noise-type label 'Weilbul' should be 'Weibull' in the first row of Table 8 and in the corresponding rows of Tables 9 and 10.
  2. [Section 4.2] The statement that the OOD entropy distribution 'shifts rightward, toward lower entropy values' is self-contradictory: a rightward shift means higher entropy, which is also what the subsequent explanation of easier detection under stronger noise implies. Please correct the direction described.
  3. [Section 3.3] The noise-injection procedure should state explicitly whether the same 20% of examples were used for the validation and test sets and across all models; the fixed seed guarantees reproducibility but not clarity about this choice.
  4. [Abstract and Introduction] The phrase 'first comprehensive comparative study' is stronger than the evidence supports, given that the literature review in Table 1 covers a small set of papers and the comparison uses a single dataset; consider tempering the wording.
  5. [Section 3.4] The number of Monte Carlo samples differs across methods (K=10 for dropout and BNN, K=4 for ensembles), and the paper justifies this as the minimum for stable results; a sentence acknowledging that entropy estimates from K=4 and K=10 may not be directly comparable would help readers interpret the differences.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the central ensemble-vs-dropout-vs-Bayesian ranking is an empirical benchmark with thresholds fit on validation and evaluated on a separate test set.

full rationale

This paper is an empirical comparative study rather than a derivation, so the classic circularity failure modes do not apply. The central claim that deep ensembles, especially De2, dominate OOD detection is measured on a held-out test set: both thresholds are computed on validation examples (tau1 by Eq. 4, tau2 by Eq. 5) and then applied to test examples in Tables 6-10, so no reported OOD or ID detection rate reduces by construction to the fitted threshold. The proposed tau2 is a validation-set F1 optimization, not a prediction derived from the test outcome. The one load-bearing self-citation, [4], justifies the ConvLSTM-D baseline architecture as effective for the same task without uncertainty; it does not determine the ensemble-versus-dropout-versus-Bayesian ranking and is therefore not load-bearing for the paper's central conclusion. The absence of a single deterministic CNN baseline is a real experimental-control concern about whether the ensemble advantage reflects the uncertainty mechanism or the underlying architectures, but it is a confound in design, not circularity as defined here. The internal inconsistency in Section 4.3 (text states deep-ensemble training is faster while Table 11 shows it is slower) is a correctness issue, not a circularity issue. No equation in the paper is equivalent to its own output by construction.

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

The study introduces no new theoretical entities or fitted model parameters beyond the decision threshold tau2 and the hand-tuned K. The main assumptions are experimental: that random splitting of overlapping windows is leakage-free, that entropy separates ID from OOD, that the architectures fairly represent each uncertainty method, and that synthetic noise matches real sensor noise.

free parameters (2)
  • K (number of Monte Carlo samples / ensemble members) = 10 for ConvLSTM-D and BNN, 4 for De1 and De2
    Hand-tuned in Section 3.2 as the minimum value for which stable results were observed; affects the entropy estimate but not the central ranking conclusion.
  • tau2 (entropy threshold based on F1-score) = varies per scenario, e.g., 0.0205 for ConvLSTM-D Scenario No.0
    Computed by maximizing F1-score on the validation set (Eq. 5). This is a threshold fitted to validation data rather than a model parameter; the paper compares it to tau1.
assumptions (4)
  • domain assumption Random split of sequentially extracted, overlapping bursts does not introduce leakage between train and test sets.
    Section 3.1: bursts of length 512 are extracted with a stride of 200 and then split 70/30 randomly. Because adjacent bursts share 312 samples, random assignment can put near-duplicates in both train and test, potentially inflating ID accuracy and OOD detection.
  • domain assumption OOD examples have higher predictive entropy than ID examples.
    Section 3.4: the entire uncertainty pipeline assumes thresholding entropy separates ID and OOD. This is the premise of the method and is not independently verified beyond the reported experiments.
  • domain assumption Each uncertainty method is fairly represented by its chosen architecture.
    Section 3.2: ConvLSTM-D, BNN, De1, and De2 use distinct network designs. The comparison assumes performance differences are attributable to the uncertainty mechanism, not the architecture.
  • domain assumption Synthetic noise injection models real sensor noise.
    Section 3.3: Gaussian, Impulse, Rayleigh, and Weibull noises are added at 20% of validation and test data. The realism of these additive noise models is assumed, not validated against field recordings.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Evaluating deep learning models for fault diagnosis of a rotating machinery with epistemic and aleatoric uncertainty." pith.science (2026). https://pith.science/paper/TEGGJ5ZC

@misc{pith2026241218980,
  author       = {Pith},
  title        = {Pith review of: Evaluating deep learning models for fault diagnosis of a rotating machinery with epistemic and aleatoric uncertainty},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TEGGJ5ZC}},
  note         = {Machine review of arXiv:2412.18980}
}
read the original abstract

Uncertainty-aware deep learning (DL) models recently gained attention in fault diagnosis as a way to promote the reliable detection of faults when out-of-distribution (OOD) data arise from unseen faults (epistemic uncertainty) or the presence of noise (aleatoric uncertainty). In this paper, we present the first comprehensive comparative study of state-of-the-art uncertainty-aware DL architectures for fault diagnosis in rotating machinery, where different scenarios affected by epistemic uncertainty and different types of aleatoric uncertainty are investigated. The selected architectures include sampling by dropout, Bayesian neural networks, and deep ensembles. Moreover, to distinguish between in-distribution and OOD data in the different scenarios two uncertainty thresholds, one of which is introduced in this paper, are alternatively applied. Our empirical findings offer guidance to practitioners and researchers who have to deploy real-world uncertainty-aware fault diagnosis systems. In particular, they reveal that, in the presence of epistemic uncertainty, all DL models are capable of effectively detecting, on average, a substantial portion of OOD data across all the scenarios. However, deep ensemble models show superior performance, independently of the uncertainty threshold used for discrimination. In the presence of aleatoric uncertainty, the noise level plays an important role. Specifically, low noise levels hinder the models' ability to effectively detect OOD data. Even in this case, however, deep ensemble models exhibit a milder degradation in performance, dominating the others. These achievements, combined with their shorter inference time, make deep ensemble architectures the preferred choice.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Generating Bearing Vibration Signals at User-Specified Fault Probabilities Using PR-GAN and Counterfactual Methods

    cs.LG 2026-07 conditional novelty 4.0 of 10

    Per-sample counterfactual optimization steers a fixed ensemble classifier's fault probability to arbitrary targets on CWRU/Paderborn vibration windows with ~0.005 error, outperforming a probability-regularized WGAN.

Reference graph

Works this paper leans on

56 extracted references · 53 canonical work pages · cited by 1 Pith paper

  1. [1]

    Y. Lei, B. Yang, X. Jiang, F. Jia, N. Li, A. K. Nandi, Applications of machine learning to machine fault diagnosis: A review and roadmap, Mechanical Systems and Signal Processing 138 (2020) 106587

    2020

  2. [2]

    R. Liu, B. Yang, E. Zio, X. Chen, Artificial intelli- gence for fault diagnosis of rotating machinery: A review, Mechanical Systems and Signal Processing 108 (2018) 33–47

    work page 2018

  3. [3]

    Zhang, S

    S. Zhang, S. Zhang, B. Wang, T. G. Habetler, Deep learning algorithms for bearing fault diagnostics—a comprehensivereview,IEEEAccess8(2020)29857– 29881

    work page 2020

  4. [4]

    M.Jalayer,C.Orsenigo,C.Vercellis, Faultdetection anddiagnosisforrotatingmachinery: Amodelbased on convolutional LSTM, fast fourier and continuous wavelet transforms, Computers in Industry 125 (2021) 103378

    work page 2021

  5. [5]

    Hoang, H.-J

    D.-T. Hoang, H.-J. Kang, A survey on deep learning based bearing fault diagnosis, Neurocomputing 335 (2019) 327–335

    work page 2019

  6. [6]

    Shojaeinasab, T

    A. Shojaeinasab, T. Charter, M. Jalayer, M. Khadivi, O. Ogunfowora, N. Raiyani, M. Yaghoubi, H. Naj- jaran, Intelligent manufacturing execution systems: A systematic review, Journal of Manufacturing Systems 62 (2022) 503–522

    work page 2022

  7. [7]

    Ovadia, E

    Y. Ovadia, E. Fertig, J. Ren, Z. Nado, D. Scul- ley, S. Nowozin, J. Dillon, B. Lakshminarayanan, 14 Reza Jalayer et al.- Preprint 1–20 J. Snoek, Can you trust your model’s uncertainty? evaluating predictive uncertainty under dataset shift, Advances in neural information processing systems 32 (2019)

    work page 2019

  8. [8]

    Abdar, F

    M. Abdar, F. Pourpanah, S. Hussain, D. Rezazade- gan, L. Liu, M. Ghavamzadeh, P. Fieguth, X. Cao, A.Khosravi,U.R.Acharya,etal., Areviewofuncer- tainty quantification in deep learning: Techniques, applications and challenges, Information Fusion 76 (2021) 243–297

    work page 2021

Show all 56 references
  1. [9]

    G. Li, L. Yang, C.-G. Lee, X. Wang, M. Rong, A Bayesian deep learning rul framework integrating epistemicandaleatoricuncertainties, IEEETransac- tionsonIndustrialElectronics68(2020)8829–8841

    2020

  2. [10]

    Y. Gal, Z. Ghahramani, Dropout as a Bayesian approximation: Representing model uncertainty in deep learning, in: international conference on ma- chine learning, PMLR, 2016, pp. 1050–1059

    2016

  3. [11]

    McClure, N

    P. McClure, N. Kriegeskorte, Representing inferen- tialuncertaintyindeepneuralnetworksthroughsam- pling,2017.URL:https://openreview.net/forum?id= HJ1JBJ5gl

    2017

  4. [12]

    A.Amini,A.Soleimany,S.Karaman,D.Rus,Spatial uncertainty sampling for end-to-end control, arXiv preprint arXiv:1805.04829 (2018)

    2018 arXiv

  5. [13]

    H. Zhou, W. Chen, L. Cheng, J. Liu, M. Xia, Trust- worthy fault diagnosis with uncertainty estimation through evidential convolutional neural networks, IEEE Transactions on Industrial Informatics 19 (2023) 10842–10852

    2023

  6. [14]

    M. Kim, K. Liu, A Bayesian deep learning frame- work for interval estimation of remaining useful life in complex systems by incorporating general degra- dation characteristics, IISE Transactions 53 (2020) 326–340

    2020

  7. [15]

    G.Mazaev,G.Crevecoeur,S.VanHoecke, Bayesian convolutional neural networks for remaining useful life prognostics of solenoid valves with uncertainty estimations, IEEE Transactions on Industrial Infor- matics 17 (2021) 8418–8428

    2021

  8. [16]

    Caceres, D

    J. Caceres, D. Gonzalez, T. Zhou, E. L. Droguett, A probabilistic Bayesian recurrent neural network for remaining useful life prognostics considering epistemic and aleatory uncertainties, Structural Control and Health Monitoring 28 (2021) e2811

    2021

  9. [17]

    W. Sun, A. R. Paiva, P. Xu, A. Sundaram, R. D. Braatz, Fault detection and identification using Bayesian recurrent neural networks, Computers & Chemical Engineering 141 (2020) 106991

    2020

  10. [18]

    S. O. Sajedi, X. Liang, Uncertainty-assisted deep visionstructuralhealthmonitoring, Computer-Aided Civil and Infrastructure Engineering 36 (2021) 126– 142

    2021

  11. [19]

    B. Qi, L. Zhang, J. Liang, J. Tong, Combinatorial techniquesforfaultdiagnosisinnuclearpowerplants based on Bayesian neural network and simplified Bayesiannetwork-artificialneuralnetwork, Frontiers in Energy Research 10 (2022) 920194

    2022

  12. [20]

    T.Zhou,L.Zhang,T.Han,E.L.Droguett,A.Mosleh, F. T. Chan, An uncertainty-informed framework for trustworthy fault diagnosis in safety-critical applica- tions, Reliability Engineering & System Safety 229 (2023) 108865

    2023

  13. [21]

    Moradi, S

    R. Moradi, S. Cofre-Martel, E. L. Droguett, M. Modarres, K. M. Groth, Integration of deep learning and Bayesian networks for condition and operation risk monitoring of complex engineering systems, Reliability Engineering & System Safety 222 (2022) 108433

    2022

  14. [22]

    T.Zhou,T.Han,E.L.Droguett, Towardstrustworthy machine fault diagnosis: A probabilistic Bayesian deep learning framework, Reliability Engineering & System Safety 224 (2022) 108525

    2022

  15. [23]

    Y. Wu, W. Jin, Y. Li, Z. Sun, J. Ren, Detecting unexpected faults of high-speed train bogie based on Bayesian deep learning, IEEE Transactions on Vehicular Technology 70 (2020) 158–172

    2020

  16. [24]

    Y. Xiao, H. Shao, M. Feng, T. Han, J. Wan, B. Liu, Towardstrustworthyrotatingmachineryfaultdiagno- sis via attention uncertainty in transformer, Journal of Manufacturing Systems 70 (2023) 186–201

    2023

  17. [25]

    Y. Feng, J. Chen, Z. Xie, J. Xie, T. Pan, C. Li, Q. Zhao, Integrating misidentification and OOD detection for reliable fault diagnosis of high-speed train bogie, IEEE Transactions on Intelligent Trans- portation Systems (2024)

    2024

  18. [26]

    L. V. Jospin, H. Laga, F. Boussaid, W. Buntine, M. Bennamoun, Hands-on Bayesian neural net- works—a tutorial for deep learning users, IEEE Computational Intelligence Magazine 17 (2022) 29– 48

    2022

  19. [27]

    Swiatkowski, K

    J. Swiatkowski, K. Roth, B. Veeling, L. Tran, J. Dil- lon, J. Snoek, S. Mandt, T. Salimans, R. Jenatton, S. Nowozin, The k-tied normal distribution: A compact parameterization of Gaussian mean field posteriors in Bayesian neural networks, in: Inter- national Conference on Mach...

    2020

  20. [28]

    Y. Gal, Z. Ghahramani, Bayesian convolutional neu- ral networks with Bernoulli approximate variational inference, arXiv preprint arXiv:1506.02158 (2015)

    2015 arXiv

  21. [29]

    D. M. Blei, A. Kucukelbir, J. D. McAuliffe, Varia- tional inference: A review for statisticians, Journal of the American statistical Association 112 (2017) 859–877

    2017

  22. [30]

    1218–1226

    T.Salimans,D.Kingma,M.Welling, MarkovChain Monte Carlo and variational inference: Bridging the gap, in: International conference on machine learning, PMLR, 2015, pp. 1218–1226

    2015

  23. [31]

    Lakshminarayanan, A

    B. Lakshminarayanan, A. Pritzel, C. Blundell, Sim- 15 Reza Jalayer et al.- Preprint 1–20 ple and scalable predictive uncertainty estimation using deep ensembles, Advances in neural informa- tion processing systems 30 (2017)

    2017

  24. [32]

    R. Hu, Q. Huang, S. Chang, H. Wang, J. He, The mbpep: a deep ensemble pruning algorithm pro- viding high quality uncertainty prediction, Applied intelligence 49 (2019) 2942–2955

    2019

  25. [33]

    S. Jain, G. Liu, J. Mueller, D. Gifford, Maximizing overall diversity for improved uncertainty estimates in deep ensembles, 2020

    2020

  26. [34]

    Zhang, B

    J. Zhang, B. Kailkhura, T. Y.-J. Han, Mix-n-match: Ensemble and compositional methods for uncer- tainty calibration in deep learning, in: International conference on machine learning, PMLR, 2020, pp. 11117–11128

    2020

  27. [35]

    Tuyet-Doan, H.-A

    V.-N. Tuyet-Doan, H.-A. Pho, B. Lee, Y.-H. Kim, Deepensemblemodelforunknownpartialdischarge diagnosis in gas-insulated switchgears using con- volutional neural networks, IEEE Access 9 (2021) 80524–80534

    2021

  28. [36]

    Han, Y.-F

    T. Han, Y.-F. Li, Out-of-distribution detection- assisted trustworthy machinery fault diagnosis ap- proach with uncertainty-aware deep ensembles, Re- liability Engineering & System Safety 226 (2022) 108648

    2022

  29. [37]

    Kafunah, M

    J. Kafunah, M. I. Ali, J. G. Breslin, Uncertainty- aware ensemble combination method for quality monitoring fault diagnosis in safety-related prod- ucts, IEEE Transactions on Industrial Informatics 20 (2023) 1975–1986

    2023

  30. [38]

    Zhang, C

    X. Zhang, C. Wang, W. Zhou, J. Xu, T. Han, Trust- worthy diagnostics with out-of-distribution detec- tion: A novel max-consistency and min-similarity guided deep ensembles for uncertainty estimation, IEEE Internet of Things Journal (2024)

    2024

  31. [39]

    Y. Mae, W. Kumagai, T. Kanamori, Uncertainty propagation for dropout-based Bayesian neural net- works, Neural Networks 144 (2021) 394–406

    2021

  32. [40]

    J.Guo,H.Zhang,D.Zhen,Z.Shi,F.Gu,A.D.Ball, An enhanced modulation signal bispectrum analysis for bearing fault detection based on non-Gaussian noisesuppression, Measurement151(2020)107240

    2020

  33. [41]

    Y.Jiang,C.Tang,X.Zhang,W.Jiao,G.Li,T.Huang, A novel rolling bearing defect detection method based on bispectrum analysis and cloud model- improved eemd, IEEE Access 8 (2020) 24323– 24333

    2020

  34. [42]

    J. Chen, G. Wang, J. Lv, Z. He, T. Yang, C. Tang, Open-set classification for signal diagnosis of ma- chinery sensor in industrial environment, IEEE Transactions on Industrial Informatics 19 (2022) 2574–2584

    2022

  35. [43]

    W.Yi,W.K.Chan,H.H.Lee,S.T.Boles,X.Zhang, An uncertainty-aware deep learning model for re- liable detection of steel wire rope defects, IEEE Transactions on Reliability (2023)

    2023

  36. [44]

    Y. Yao, T. Han, J. Yu, M. Xie, Uncertainty-aware deeplearningforreliablehealthmonitoringinsafety- critical energy systems, Energy 291 (2024) 130419

    2024

  37. [45]

    D. Li, Y. Wang, J. Wang, C. Wang, Y. Duan, Recent advancesinsensorfaultdiagnosis: Areview,Sensors and Actuators A: Physical 309 (2020) 111990

    2020

  38. [46]

    M.Qiao,S.Yan,X.Tang,C.Xu, Deepconvolutional andLSTMrecurrentneuralnetworksforrollingbear- ing fault diagnosis under strong noises and variable loads, IEEE Access 8 (2020) 66257–66269

    2020

  39. [47]

    Cheng, L

    W. Cheng, L. Zhang, J. Xing, X. Chen, Z. Nie, S. Zhang, S. Wang, R. Zhang, Q. Huang, Three- types-of-graph-relational guided domain adaptation approach for fault diagnosis of nuclear power circu- lating water pump, IEEE Transactions on Industrial Informatics 20 (2023) 1348–1359

    2023

  40. [48]

    Salakhutdinov, Dropout: a simple way to prevent neural networks from overfitting, The journal of machine learning research 15 (2014) 1929–1958

    N.Srivastava,G.Hinton,A.Krizhevsky,I.Sutskever, R. Salakhutdinov, Dropout: a simple way to prevent neural networks from overfitting, The journal of machine learning research 15 (2014) 1929–1958

    2014

  41. [49]

    Neupane, J

    D. Neupane, J. Seok, Bearing fault detection and diagnosis using Case Western Reserve University dataset with deep learning approaches: A review, IEEE Access 8 (2020) 93155–93178

    2020

  42. [50]

    Y. Lu, Z. Song, Q. Gao, D. Zhu, D. Sun, Bearing fault diagnosis based on multi-band filtering, IET Science, Measurement & Technology 16 (2022) 101–117

    2022

  43. [51]

    A.Moshrefzadeh,A.Fasana,J.Antoni, Thespectral amplitude modulation: A nonlinear filtering pro- cess for diagnosis of rolling element bearings, Me- chanical Systems and Signal Processing 132 (2019) 253–276

    2019

  44. [52]

    Hebda-Sobkowicz, R

    J. Hebda-Sobkowicz, R. Zimroz, A. Wyłomańska, J. Antoni, Infogram performance analysis and its enhancement for bearings diagnostics in presence of non-Gaussian noise, Mechanical Systems and Signal Processing 170 (2022) 108764

    2022

  45. [53]

    J. Peng, Y. Zhao, X. Zhang, J. Wang, L. Wang, An adaptive reweighted-kurtogram for bearing fault diagnosis under strong external impulse noise, Structural Health Monitoring (2024) 14759217231226267

    2024

  46. [54]

    X. Du, L. Jia, I. U. Haq, Fault diagnosis based on SPBO-SDAE and transformer neural network for rotating machinery, Measurement 188 (2022) 110545

    2022

  47. [55]

    Lessmeier, J

    C. Lessmeier, J. K. Kimotho, D. Zimmer, W. Sextro, Condition monitoring of bearing damage in elec- tromechanical drive systems by using motor current signals of electric motors: A benchmark data set for data-driven classification, 2016

    2016

  48. [56]

    same" Batch Normalization (Convolutional layer)

    M. Sehri, P. Dumond, M. Bouchard, University of Ottawaconstantloadandspeedrolling-elementbear- 16 Reza Jalayer et al.- Preprint 1–20 Table A1 Hyperparameter of the ConvLSTM-D architecture. Input layer Number of nodes 512 1st Conv + Pool Activation function (Convolutional layer...

    2023

Pith tools

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