Pith. sign in

REVIEW 5 major objections 5 minor 52 references

Toward Decision-Oriented Prognostics: An Integrated Estimate-Optimize Framework for Predictive Maintenance

T0 review · 5 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Training a remaining-useful-life model directly on expected maintenance cost, rather than on prediction error, reduces average maintenance regret by up to about 22% relative to the standard estimate-then-optimize approach in the paper's…

desk verdict A legitimate application of decision-focused learning to RUL-based maintenance, but the headline quantile-policy gain may largely be a reliability-constraint violation, and the theory covers a different algorithm than the one actually run. read the letter →

arxiv 2506.19698 v1 pith:PBC6EZTN submitted 2025-06-24 cs.AI math.OC

Zhuojun Xie , Adam Abdin , Yiping Fang This is my paper · ORCID
classification cs.AImath.OC
keywords MachineLearningPredictiveMaintenanceEstimate-then-OptimizeIntegratedEstimate-Optimizeremainingusefullifedecision-focusedmodelmisspecificationC-MAPSSturbofan
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 claims that how well a remaining-useful-life (RUL) model predicts is not the same as how well it supports maintenance decisions: under model misspecification a more accurate forecast can produce costlier maintenance, because the discrete set of maintenance windows makes decision cost jump discontinuously with small prediction errors. To fix this, the authors propose an integrated estimate-optimize (IEO) framework that starts from a forecast model trained on prediction error and fine-tunes it directly on expected maintenance cost, using a stochastic perturbation gradient method to handle the non-differentiability of discrete decisions. They prove a finite-sample consistency guarantee: with enough data, minimizing the decision loss drives excess decision risk toward zero at rate $O(\sqrt{\log n/n})$, covering both stochastic-optimization and quantile maintenance policies. On C-MAPSS turbofan data they report that IEO reduces average maintenance regret by up to about 22% relative to estimate-then-optimize, with the gains persisting as misspecification intensifies. The practical point: prognostic models should be tuned for the decisions they feed, not for the accuracy of their predictions alone.

What carries the argument

The central object is the decision loss $L(m(X;\omega),Y) = c(\pi(m(X;\omega);\phi),Y)$, which scores a prognostic model by the maintenance cost of the decision its predicted distribution triggers; it is consistent with the decision objective by construction, so minimizing it is equivalent to minimizing expected cost. The model produces a Weibull-type distribution for the remaining useful life, outputting just two parameters (scale $\lambda$ and shape $k$) that are truncated and renormalized onto a discrete support, keeping the prediction low-dimensional while retaining flexibility. Because the policy's argmin over discrete maintenance windows is not differentiable, the gradient of the decision loss with respect to the Weibull parameters is estimated by a stochastic perturbation scheme: Gaussian noise is added to the parameters and the score-function identity $\nabla_\theta \tilde{L}_i = \Sigma^{-1}\mathbb{E}_\eta[L(\theta_i+\Sigma\eta,y_i)\eta]$ is evaluated by Monte-Carlo sampling, with a REINFORCE-style baseline to reduce variance. The generalization argument rests on the Natarajan dimension $d$ of the integrated policy class, which enters a Sauer-Massart bound to yield the finite-sample excess-risk guarantee.

What would settle it

Replicate the long-term experiment on the FD003 subset of C-MAPSS (two fault modes, same sliding-window features, Weibull head, 200-step ETO pretraining plus 100-step IEO fine-tuning, RUL capped at 125, 20 test engines, 100 repetitions): if the integrated fine-tuning does not reduce average maintenance regret below estimate-then-optimize in this independently generated misspecification regime, the claimed structural advantage of decision-loss training fails to generalize beyond the single-mode FD001 setting.

Watch

Extended reading notes

Core claim

The paper's central claim is that prediction accuracy and decision quality are different objectives in predictive maintenance, and that training should target the latter. The authors first show that the optimality gap of a maintenance policy is not Lipschitz continuous in the estimation error of the RUL distribution, so two estimates with identical accuracy can lead to different expected costs and a worse prediction can lead to better decisions. They then define the decision loss $L(m(X;\omega),Y) = c(\pi(m(X;\omega);\phi),Y)$, which scores a model by the maintenance cost of the decision its predicted distribution triggers, and note that this loss is consistent with the decision objective by construction. The main theoretical result is a PAC-type guarantee: an empirical minimizer of this decision loss has excess decision risk bounded by $2C_1\sqrt{\log(2/\delta)/2n} + 2C_1\sqrt{(2d\log(en/d)+4d\log K)/n}$, where $d$ is the Natarajan dimension of the integrated policy class and $K$ the number of maintenance windows, so the learned model converges to the best-in-class decision-oriented model as the sample size grows. Empirically, on the C-MAPSS FD001 turbofan data with a Weibull-type RUL model, IEO fine-tuning reduces average regret by 4.3% (CSO) and 27.8% (quantile) in the in-sample base case, and by 6.56% and 21.92% respectively in the long-term out-of-sample setting under strong misspecification.

Load-bearing premise

The paper's guarantee applies to a model that exactly minimizes the decision loss on the training data and presupposes that the capacity measure of the model class stays below the sample size, while the experiments use a 100-step fine-tune and never compute that capacity measure, so the proof covers an idealized training procedure rather than the one whose regret reductions are reported.

Editorial extensions

If this is right

  • Prognostic models should be selected and tuned on maintenance cost: in the reported experiments, the model with the lowest prediction error is not the model with the lowest maintenance regret, so accuracy-based model comparison can select the wrong model for operations.
  • The advantage of decision-oriented training persists as misspecification intensifies: for the stochastic-optimization policy the relative regret reduction rises from 4.3% in the base case to 6.56% in the long-term setting, and for the quantile policy worst-case regret over 100 trials drops from 51.1 to 30.8.
  • The quantile policy benefits most when its risk tolerance is misaligned with the cost structure: with a fixed tolerance, IEO corrects the over-conservatism that inflates regret, calibrating the observed failure frequency toward the preset risk margin.
  • The finite-sample guarantee means the approach is usable on the small run-to-failure datasets typical of maintenance practice, since excess decision risk vanishes at rate $O(\sqrt{\log n/n})$ as the number of run-to-failure samples grows.
  • IEO stabilizes decisions across training runs: maximum regret over 100 trials falls by 12.9% (CSO) and 34.1% (quantile) in the base case, so decision-oriented fine-tuning also reduces run-to-run variability.

Reading between the lines

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

  • The non-Lipschitz mechanism behind Proposition 1 is not specific to maintenance: any forecast that feeds a discrete choice problem (inventory orders, scheduling slots, capacity decisions) can exhibit the same divergence between prediction accuracy and decision quality, so the decision-loss fine-tuning recipe should transfer to those settings.
  • The sharp drop in regret variance suggests that decision-oriented fine-tuning also acts as a regularizer, pulling different random initializations toward the same decision regions; a direct test would be to measure the variance of the chosen maintenance windows, not just costs, across seeds.
  • Because the paper's own Remark 1 concedes that sliding-window time series violate the i.i.d. assumption behind Theorem 1, the theoretical guarantee is not literally in force for the experiments; re-deriving the bound under a mixing assumption and checking the regret reduction under block-bootstrap resampling would close that gap.
  • The Weibull head can only represent unimodal RUL distributions, leaving multi-modal degradation (the two-fault-mode data the paper sets aside) untested; a mixture-of-Weibull extension, which the paper names as future work, would show whether decision-loss training can compensate for that restriction.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 5 minor

Summary. The paper proposes an integrated estimate-optimize (IEO) framework for predictive maintenance, in which a probabilistic remaining-useful-life model is fine-tuned on the expected maintenance cost rather than on a purely predictive loss. The authors contrast this with the estimate-then-optimize (ETO) baseline, provide a formal consistency analysis via a non-asymptotic generalization bound in Theorem 1 and Corollary 1, and introduce a stochastic perturbation gradient estimator to handle the non-differentiability of discrete maintenance decisions. Experiments on the CMAPSS turbofan dataset are run across three settings: an in-sample base case, a short-term out-of-sample case, and a long-term out-of-sample case with artificially induced stronger misspecification. The paper reports average regret reductions up to 21.92% for the quantile policy and 6.56% for the contextual stochastic optimization policy in the long-term experiment.

Significance. If the central claims hold, the paper makes a useful contribution to decision-focused learning and predictive maintenance by connecting statistical consistency of prognostic models to downstream maintenance costs. The strengths include a clearly formulated empirical decision-risk minimization problem, an unbiased score-function gradient estimator in Eq. (13) that avoids differentiating through a discrete argmin, a non-asymptotic bound built on Natarajan dimension, and an experimental design that varies the degree of misspecification and repeats each experiment 100 times. The main novelty claim, namely, adapting IEO to discrete maintenance decision spaces, is plausible and relevant. However, the significance is currently moderated by a gap between the proven guarantee and the implemented algorithm, and by a headline empirical comparison in which the quantile policy violates its own reliability constraint. These issues are addressable, so the contribution is worth pursuing in a major revision.

major comments (5)
  1. [Section 4.2, Eq. (11), and Table 1] Corollary 1 is stated for omega_erm, an exact minimizer of the empirical decision risk bRL(omega), but the implemented IEO procedure runs only 100 steps of Adam from the ETO-trained parameter omega_eto. The paper gives no convergence guarantee for this fine-tuning scheme and no experiment measuring how close the 100-step iterate is to omega_erm. As written, the finite-sample guarantee therefore applies to a quantity that the deployed algorithm is not shown to produce. The authors should either prove a convergence result for the fine-tuning procedure, report the distance between the fine-tuned iterate and a candidate empirical minimizer, or explicitly state that Theorem 1 and Corollary 1 characterize exact ERM rather than the 100-step algorithm.
  2. [Section 5.4, Table 4, and the regret metric (i) in Section 3.4] The headline claim of a 21.92% regret reduction for IEO-Q is not an apples-to-apples comparison. The quantile policy in Eq. (4) is defined by a target failure probability alpha=0.01, but IEO-Q attains a realized failure frequency of 0.025 +/- 0.008 in the long-term experiment, about 2.5 times the target, whereas ETO-Q is over-conservative at 0.001 +/- 0.002. The regret metric (i) does not penalize violation of the alpha constraint, so part of IEO-Q's cost reduction may come from shifting preventive maintenance cost into occasional failures. If alpha is a hard reliability requirement, IEO-Q is inadmissible in this experiment and the 21.92% figure overstates the framework's benefit. The authors should compare policies at matched realized failure frequencies, add a constraint-aware metric with a penalty for alpha violation, or clearly state whether alpha is a hard constraint or merely a tuning parameter.
  3. [Section 2.3 and Appendix A] Proposition 1 asserts that the relative optimality gap delta_o is not Lipschitz continuous in the estimation error delta_e and states that this holds for all C>0. The proposition is presented as a formal statement but no proof is provided in the main text or in Appendix A, which contains only the proofs of Theorem 1 and Corollary 1. Since the motivating example is described as being formally captured by this proposition, a complete proof or a precise construction of the distributions and cost parameters realizing the failure of Lipschitz continuity should be supplied.
  4. [Theorem 1, Definition 4, and Remark 1] The non-asymptotic bound is stated in terms of the Natarajan dimension d of the integrated policy class G and assumes d <= n, but d is never computed or bounded for the Weibull-plus-argmin neural network used in Section 5. Consequently, the bound is not instantiated for the actual model class deployed in the experiments, and it is unclear whether the regime d <= n holds there. In addition, Remark 1 concedes that the i.i.d. assumption underlying Assumption 1 is likely violated by the sliding-window construction, and no dependent-data guarantee is derived. The paper should either compute or upper-bound d for the implemented architecture, or present Theorem 1 explicitly as an abstract existence result and supply additional analysis or experiments for the dependent-data setting.
  5. [Section 5.4] The long-term experiment is described as introducing a 'two-phase degradation process' by capping RUL values greater than 125 at 125. This is a manual label transformation rather than a two-phase degradation model, and the resulting 'stronger misspecification' is an artifact of the capping rule. The interpretation of the comparison would be clearer if the paper stated this explicitly and discussed how the capping changes the label distribution relative to the original FD001 test set, which is not used in the out-of-sample experiments.
minor comments (5)
  1. [Section 5.1] The list of framework variants says 'ETO-C, ETO-C, IEO-C, and IEO-Q'; the second entry should presumably be ETO-Q.
  2. [Section 2.3] The sentence 'The smaller delta_e(bP), the better decision is incurred by bP' is garbled; it should be 'the smaller delta_o(bP), the better the decision incurred by bP'.
  3. [Section 3.2, Proposition 2] The proof of Proposition 2 is definitional because RL(omega)=R(omega) by construction. The authors should state explicitly that this is an identity rather than a substantive consistency result, otherwise the proposition reads as circular.
  4. [Section 5.2, Table 2] The phrase 'well-calibrated failure frequency, closely matching the preset risk margin of 1%' is accurate for the base case but should not be generalized, since the long-term experiment in Table 4 shows a failure frequency of 0.025, which is 2.5 times the target.
  5. [Section 5.3 and Section 5.4] The out-of-sample experiments use held-out engines from the FD001 training set with an added cap on RUL labels, rather than the original CMAPSS test set. The paper should state this limitation clearly and discuss whether the conclusions are expected to transfer to the original test distribution.

Circularity Check

1 steps flagged · score 2.0 of 10

One definitional tautology in Proposition 2; the central PAC bound and the out-of-sample experiments are not circular.

  1. self definitional [Section 3.2, Proposition 2, Equations (8) and (10)]
    "We show the loss L holds the P-consistency by the following Proposition 2. Proposition 2. The loss function L in Equation (10) is P-consistent for any model class M, since by definition RL(ω) = R(ω)."

    The decision risk R(ω) in Equation (8) is defined as E_P[c(π(m(X;ω);ϕ);Y)], and the proposed loss L in Equation (10) is defined as c(π(m(X);ϕ);Y). Therefore RL(ω) = R(ω) holds by direct substitution, not by any argument. Proposition 2 then restates that a near-minimizer of the decision loss is a near-minimizer of the decision risk, which is immediate because the two risks are the same function. This is a definitional tautology rather than a substantive consistency result for a surrogate loss. It is not load-bearing for the main finite-sample guarantee: Theorem 1 and Corollary 1 are proven from the Natarajan-dimension Rademacher bound and McDiarmid/Hoeffding arguments independently of this identity.

full rationale

The paper's central theoretical claim is Corollary 1, a non-asymptotic generalization bound for empirical risk minimization of the decision loss L. That proof uses Assumptions 1-3, the Natarajan dimension of the integrated policy class, Sauer's lemma, Rademacher complexity, and Hoeffding/McDiarmid inequalities; it does not reduce to its own input. No load-bearing self-citation chain is present, and no uniqueness theorem by the authors is invoked to forbid alternatives. The empirical headline comparison in Table 4 is out-of-sample: the model is fine-tuned on 80 engines and evaluated on 20 held-out engines, so the reported 21.92% regret reduction is not simply a training-set fit. The base-case experiment is explicitly labeled in-sample and should not be read as an out-of-sample prediction; because it is presented as an ideal-case diagnostic rather than as the main generalization evidence, it does not by itself make the paper circular. The separate concern that IEO-Q violates the alpha=1% reliability target while ETO-Q is over-conservative is a question of whether the regret metric incorporates a hard constraint, which is an evaluation-fairness or correctness issue, not a circularity of the derivation. The only clear definitional reduction is Proposition 2, whose P-consistency claim is true by construction because L is defined as the decision cost itself. Since that proposition is acknowledged in the text and is not needed for the generalization bound, it warrants a low score rather than a charge that the paper's central result is forced.

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

The central claim rests on a standard supervised learning setup plus a set of hand-chosen hyperparameters (quantile margin alpha=0.01, perturbation covariance I, M=1000, support size H=150, and cost coefficients) that shape the experimental results. The theory further assumes i.i.d. data, a finite decision set, a bounded cost, and an exact empirical risk minimizer, none of which perfectly matches the implemented algorithm or the sliding-window time-series data. The most fragile unpaid premise is the assumption that the Natarajan dimension of the integrated policy class is small enough for the stated bound to be meaningful.

free parameters (5)
  • Quantile risk margin alpha = 0.01
    Hand-set tolerance for the quantile policy; directly controls the safety constraint and influences how much IEO can reduce regret by overriding conservatism.
  • Perturbation covariance Sigma = Identity matrix I
    Chosen as identity in Section 5.1; determines the smoothness and variance of the stochastic perturbation gradient estimate.
  • Monte Carlo samples M = 1000
    Number of noise samples for each gradient estimate; larger M reduces variance but increases compute; hand-set.
  • RUL support size H = 150
    Number of discrete RUL intervals; affects resolution and complexity of the classification model.
  • Maintenance cost coefficients (cp, cc, cm, cd) = cp=50, cc=200, cm=1, cd=5
    Cost parameters for the turbofan experiments; they define the downstream objective and determine the magnitude of regret and the degree of policy misalignment.
assumptions (5)
  • domain assumption Dataset is i.i.d. from the data-generating process (Assumption 1)
    Needed for the Rademacher/McDiarmid bounds in Theorem 1; Remark 1 concedes sliding-window time-series data likely violate this.
  • domain assumption Maintenance cost function is bounded by C1 (Assumption 2)
    Required for Massart's lemma and Hoeffding's inequality; holds when Y and Z are finite.
  • domain assumption Feasible decision set Z is finite (Assumption 3)
    Enables use of Natarajan dimension and finite-class growth function bounds.
  • ad hoc to paper Natarajan dimension d of integrated policy class G satisfies d <= n
    Theorem 1 and Corollary 1 assume d <= n and use it to apply Sauer's lemma; the paper never bounds or estimates d for its neural network policy class.
  • ad hoc to paper Existence of an exact empirical risk minimizer omega_erm (Corollary 1)
    The theory applies to a minimizer of the empirical decision risk, but the implemented algorithm only performs 100 fine-tuning steps and does not solve the minimization.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Toward Decision-Oriented Prognostics: An Integrated Estimate-Optimize Framework for Predictive Maintenance." pith.science (2026). https://pith.science/paper/PBC6EZTN

@misc{pith2026250619698,
  author       = {Pith},
  title        = {Pith review of: Toward Decision-Oriented Prognostics: An Integrated Estimate-Optimize Framework for Predictive Maintenance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PBC6EZTN}},
  note         = {Machine review of arXiv:2506.19698}
}
read the original abstract

Recent research increasingly integrates machine learning (ML) into predictive maintenance (PdM) to reduce operational and maintenance costs in data-rich operational settings. However, uncertainty due to model misspecification continues to limit widespread industrial adoption. This paper proposes a PdM framework in which sensor-driven prognostics inform decision-making under economic trade-offs within a finite decision space. We investigate two key questions: (1) Does higher predictive accuracy necessarily lead to better maintenance decisions? (2) If not, how can the impact of prediction errors on downstream maintenance decisions be mitigated? We first demonstrate that in the traditional estimate-then-optimize (ETO) framework, errors in probabilistic prediction can result in inconsistent and suboptimal maintenance decisions. To address this, we propose an integrated estimate-optimize (IEO) framework that jointly tunes predictive models while directly optimizing for maintenance outcomes. We establish theoretical finite-sample guarantees on decision consistency under standard assumptions. Specifically, we develop a stochastic perturbation gradient descent algorithm suitable for small run-to-failure datasets. Empirical evaluations on a turbofan maintenance case study show that the IEO framework reduces average maintenance regret up to 22% compared to ETO. This study provides a principled approach to managing prediction errors in data-driven PdM. By aligning prognostic model training with maintenance objectives, the IEO framework improves robustness under model misspecification and improves decision quality. The improvement is particularly pronounced when the decision-making policy is misaligned with the decision-maker's target. These findings support more reliable maintenance planning in uncertain operational environments.

Figures

Figures reproduced from arXiv: 2506.19698 by the authors.

Figure 1
Figure 1. Estimated probability and relative optimality gap in the example. Note: [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Gradients generated by the ETO and the IEO approaches. b [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗
Figure 3
Figure 3. Performance of four frameworks under different evaluation metrics in the ideal case. [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Performance of four frameworks under different evaluation metrics in the short-term scheduling [PITH_FULL_IMAGE:figures/full_fig_p016_4.png]
Figure 5
Figure 5. Figure 5: Performance of four frameworks under different evaluation metrics in the long-term scheduling [PITH_FULL_IMAGE:figures/full_fig_p017_5.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 49 canonical work pages

  1. [1]

    Atamuradov, Vepa, Kamal Medjaher, Pierre Dersin, Benjamin Lamoureux, Noureddine Zerhouni. 2017. Prognostics and health management for maintenance practitioners-review, implementation and tools evaluation. International Journal of Prognostics and Health Management\/ 8 (3) 1--31

    work page 2017

  2. [2]

    Ban, Gah-Yi, Cynthia Rudin. 2019. The big data newsvendor: Practical insights from machine learning. Operations Research\/ 67 (1) 90--108

    work page 2019

  3. [3]

    Bartlett, Peter L, Andrea Montanari, Alexander Rakhlin. 2021. Deep learning: a statistical viewpoint. Acta numerica\/ 30 87--201

    work page 2021

  4. [4]

    Basciftci, Beste, Shabbir Ahmed, Nagi Gebraeel. 2020. Data-driven maintenance and operations scheduling in power systems under decision-dependent uncertainty. IISE transactions\/ 52 (6) 589--602

    work page 2020

  5. [5]

    Berthet, Quentin, Mathieu Blondel, Olivier Teboul, Marco Cuturi, Jean-Philippe Vert, Francis Bach. 2020. Learning with differentiable pertubed optimizers. Advances in Neural Information Processing Systems\/ 33 9508--9519

    work page 2020

  6. [6]

    Bertsimas, Dimitris, Nathan Kallus. 2020. From predictive to prescriptive analytics. Management Science\/ 66 (3) 1025--1044

    work page 2020

  7. [7]

    Cai, Yue, Ruud H Teunter, Bram de Jonge. 2023. A data-driven approach for condition-based maintenance optimization. European Journal of Operational Research\/ 311 (2) 730--738

    work page 2023

  8. [8]

    Canavos, George C, Chris P Taokas. 1973. Bayesian estimation of life parameters in the weibull distribution. Operations Research\/ 21 (3) 755--763

    work page 1973

Show all 52 references
  1. [9]

    De Jonge, Bram, Philip A Scarf. 2020. A review on maintenance optimization. European Journal of Operational Research\/ 285 (3) 805--824

    2020

  2. [10]

    Drent, Collin, Melvin Drent, Joachim Arts. 2024. Condition-based production for stochastically deteriorating systems: optimal policies and learning. Manufacturing & Service Operations Management\/ 26 (3) 1137--1156

    2024

  3. [11]

    Drent, Collin, Melvin Drent, Joachim Arts, Stella Kapodistria. 2023. Real-time integrated learning and decision making for cumulative shock degradation. Manufacturing & Service Operations Management\/ 25 (1) 235--253

    2023

  4. [12]

    El Balghiti, Othman, Adam N Elmachtoub, Paul Grigas, Ambuj Tewari. 2023. Generalization bounds in the predict-then-optimize framework. Math. Oper. Res.\/ 48 (4) 2043--2065

    2023

  5. [13]

    predict, then optimize

    Elmachtoub, Adam N, Paul Grigas. 2022. Smart “predict, then optimize”. Management Science\/ 68 (1) 9--26

    2022

  6. [14]

    Elmachtoub, Adam N., Henry Lam, Haixiang Lan, Haofeng Zhang. 2025. Dissecting the impact of model misspecification in data-driven optimization. The 28th International Conference on Artificial Intelligence and Statistics\/

    2025

  7. [15]

    Elmachtoub, Adam N, Henry Lam, Haofeng Zhang, Yunfan Zhao. 2023. Estimate-then-optimize versus integrated-estimation-optimization versus sample average approximation: a stochastic dominance perspective. arXiv preprint arXiv:2304.06833\/

    2023 arXiv

  8. [16]

    Gebraeel, Nagi, Yaguo Lei, Naipeng Li, Xiaosheng Si, Enrico Zio, et al. 2023. Prognostics and remaining useful life prediction of machinery: advances, opportunities and challenges. Journal of Dynamics, Monitoring and Diagnostics\/ 1--12

    2023

  9. [17]

    Haussler, David, Philip M Long. 1995. A generalization of sauer's lemma. Journal of Combinatorial Theory, Series A\/ 71 (2) 219--240

    1995

  10. [18]

    Ho-Nguyen, Nam, Fatma K l n c -Karzan. 2022. Risk guarantees for end-to-end prediction and optimization processes. Management Science\/ 68 (12) 8680--8698

    2022

  11. [19]

    Hu, Yichun, Nathan Kallus, Xiaojie Mao. 2022. Fast rates for contextual linear optimization. Management Science\/ 68 (6) 4236--4245

    2022

  12. [20]

    Kamariotis, Antonios, Konstantinos Tatsis, Eleni Chatzi, Kai Goebel, Daniel Straub. 2024. A metric for assessing and optimizing data-driven prognostic algorithms for predictive maintenance. Reliability Engineering & System Safety\/ 242 109723

    2024

  13. [21]

    Kannan, Rohit, G \"u zin Bayraksan, James R Luedtke. 2025. Data-driven sample average approximation with covariate information. Operations Research\/

    2025

  14. [22]

    Kingma, Diederik P, Jimmy Ba. 2014. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980\/

    2014 arXiv

  15. [23]

    Lagos, Carlos, Felipe Delgado, Mathias A Klapp. 2020. Dynamic optimization for airline maintenance operations. Transportation Science\/ 54 (4) 998--1015

    2020

  16. [24]

    Lee, Hyunju, Ji Hwan Cha. 2016. New stochastic models for preventive maintenance and maintenance optimization. European Journal of Operational Research\/ 255 (1) 80--90

    2016

  17. [25]

    Lee, Juseong, Mihaela Mitici. 2023. Deep reinforcement learning for predictive aircraft maintenance using probabilistic remaining-useful-life prognostics. Reliability Engineering & System Safety\/ 230 108908

    2023

  18. [26]

    Lei, Yaguo, Naipeng Li, Liang Guo, Ningbo Li, Tao Yan, Jing Lin. 2018. Machinery health prognostics: A systematic review from data acquisition to rul prediction. Mechanical systems and signal processing\/ 104 799--834

    2018

  19. [27]

    Lewis, Austin D, Katrina M Groth. 2022. Metrics for evaluating the performance of complex engineering system health monitoring models. Reliability Engineering & System Safety\/ 223 108473

    2022

  20. [28]

    Li, Zhe, Qian He, Jingyue Li. 2024. A survey of deep learning-driven architecture for predictive maintenance. Engineering applications of artificial intelligence\/ 133 108285

    2024

  21. [29]

    Lin, Shaochong, Youhua Chen, Yanzhi Li, Zuo-Jun Max Shen. 2022. Data-driven newsvendor problems regularized by a profit risk constraint. Production and Operations Management\/ 31 (4) 1630--1644

    2022

  22. [30]

    Lv, Jinchi, Jun S Liu. 2014. Model selection principles in misspecified models. Journal of the Royal Statistical Society Series B: Statistical Methodology\/ 76 (1) 141--167

    2014

  23. [31]

    Mandi, Jayanta, James Kotary, Senne Berden, Maxime Mulamba, Victor Bucarey, Tias Guns, Ferdinando Fioretto. 2024. Decision-focused learning: Foundations, state of the art, benchmark and future opportunities. Journal of Artificial Intelligence Research\/ 80 1623--1701

    2024

  24. [32]

    Mitici, Mihaela, Ingeborg de Pater, Anne Barros, Zhiguo Zeng. 2023. Dynamic predictive maintenance for multiple components using data-driven probabilistic rul prognostics: The case of turbofan engines. Reliability Engineering & System Safety\/ 234 109199

    2023

  25. [33]

    Mohamed, Shakir, Mihaela Rosca, Michael Figurnov, Andriy Mnih. 2020. Monte carlo gradient estimation in machine learning. Journal of Machine Learning Research\/ 21 (132) 1--62

    2020

  26. [34]

    Mohri, Mehryar, Afshin Rostamizadeh, Ameet Talwalkar. 2018. Foundations of Machine Learning\/ . 2nd ed. MIT Press, Cambridge, MA

    2018

  27. [35]

    Nguyen, Khanh TP, Kamal Medjaher. 2019. A new dynamic predictive maintenance framework using deep learning for failure prognostics. Reliability Engineering & System Safety\/ 188 251--262

    2019

  28. [36]

    Papamakarios, George, Eric Nalisnick, Danilo Jimenez Rezende, Shakir Mohamed, Balaji Lakshminarayanan. 2021. Normalizing flows for probabilistic modeling and inference. Journal of Machine Learning Research\/ 22 (57) 1--64

    2021

  29. [37]

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

    2019

  30. [38]

    Pintelon, Liliane M, LF Gelders. 1992. Maintenance management decision making. European Journal of Operational Research\/ 58 (3) 301--317

    1992

  31. [39]

    Qi, Meng, Paul Grigas, Zuo-Jun Max Shen. 2021. Integrated conditional estimation-optimization. arXiv preprint arXiv:2110.12351\/

    2021 arXiv

  32. [40]

    Qi, Meng, Yuanyuan Shi, Yongzhi Qi, Chenxin Ma, Rong Yuan, Di Wu, Zuo-Jun Shen. 2023. A practical end-to-end inventory management model with deep learning. Management Science\/ 69 (2) 759--773

    2023

  33. [41]

    Sadana, Utsav, Abhilash Chenreddy, Erick Delage, Alexandre Forel, Emma Frejinger, Thibaut Vidal. 2025. A survey of contextual optimization methods for decision-making under uncertainty. European Journal of Operational Research\/ 320 (2) 271--289

    2025

  34. [42]

    Saxena, Abhinav, Kai Goebel, Don Simon, Neil Eklund. 2008. Damage propagation modeling for aircraft engine run-to-failure simulation. 2008 international conference on prognostics and health management\/ . IEEE, 1--9

    2008

  35. [43]

    Shalev-Shwartz, Shai, Shai Ben-David. 2014. Understanding machine learning: From theory to algorithms\/ . Cambridge university press

    2014

  36. [44]

    Sim, Melvyn, Qinshen Tang, Minglong Zhou, Taozeng Zhu. 2024. The analytics of robust satisficing: predict, optimize, satisfice, then fortify. Operations Research\/

    2024

  37. [45]

    Srivastava, Nitish, Geoffrey Hinton, Alex Krizhevsky, Ilya Sutskever, Ruslan Salakhutdinov. 2014. Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res.\/ 15 (1) 1929–1958

    2014

  38. [46]

    Stratigakos, Akylas, Salvador Pineda, Juan Miguel Morales. 2024. Decision-focused linear pooling for probabilistic forecast combination. International Journal of Forecasting\/

    2024

  39. [47]

    van der Haar, Joost F, Collin Drent, Robert N Boute. 2025. Predict-and-optimize machine learning for feature-based maintenance optimization. Available at SSRN 5225643\/

    2025

  40. [48]

    Wagenaar, Joris C, Leo G Kroon, Marie Schmidt. 2017. Maintenance appointments in railway rolling stock rescheduling. Transportation Science\/ 51 (4) 1138--1160

    2017

  41. [49]

    Waldhauser, Felix, Hamza Boukabache, Daniel Perrin, Martin Dazer. 2024. Generating realistic failure data for predictive maintenance: A simulation and cgan-based methodology. PHM Society European Conference\/ 8 (1) 4

    2024

  42. [50]

    Wen, Yuxin, Md Fashiar Rahman, Honglun Xu, Tzu-Liang Bill Tseng. 2022. Recent advances and trends of predictive maintenance from data-driven machine prognostics perspective. Measurement\/ 187 110276

    2022

  43. [51]

    Williams, Ronald J. 1992. Simple statistical gradient-following algorithms for connectionist reinforcement learning. Machine learning\/ 8 229--256

    1992

  44. [52]

    Zhuang, Liangliang, Ancha Xu, Xiao-Lin Wang. 2023. A prognostic driven predictive maintenance framework based on bayesian deep learning. Reliability Engineering & System Safety\/ 234 109181

    2023

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.