Pith. sign in

REVIEW 4 major objections 2 minor 2 cited by

VAE-DNN: Energy-Efficient Trainable-by-Parts Surrogate Model For Parametric Partial Differential Equations

T0 review · 4 major / 2 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A surrogate model trained piece-by-piece beats FNO and DeepONet on a groundwater equation.

desk verdict The submitted manuscript is two unrelated papers stapled together; the VAE-DNN claims have no supporting text, so nothing can be verified. read the letter →

arxiv 2508.03839 v1 pith:TOZJ4OPV submitted 2025-08-05 cs.LG cs.AIcs.CE

Yifei Zong , Alexandre M. Tartakovsky This is my paper
classification cs.LGcs.AIcs.CE MSC 68T0765M99
keywords trainable-by-partssurrogatemodelvariationalautoencoderoperatorlearningforwardinverseproblemsnonlineardiffusionequationgroundwaterflowenergy-efficienttraining
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes a surrogate model, VAE-DNN, for solving parametric nonlinear PDEs that can be trained in three independent parts. The encoder is trained as part of a variational autoencoder for the input field, the decoder as part of a VAE for the solution field, and a fully connected network maps one latent space to the other. The authors claim that on the nonlinear diffusion equation describing groundwater flow in an unconfined aquifer, VAE-DNN attains higher accuracy while consuming less training time and energy than the FNO and DeepONet baselines, for both forward and inverse problems. If correct, this makes high-fidelity surrogates substantially cheaper to build.

What carries the argument

The key object is the two-VAE architecture: a variational autoencoder maps the input field $y(\bm{x})$ to a latent code $\bm{\mu}_{\phi_y}$, a separate variational autoencoder maps the solution field $h(\bm{x},t)$ to $\bm{\mu}_{\phi_h}$, and a fully connected neural network is trained to map between the two latent codes. Because the VAEs are trained on their respective fields alone, the encoder and decoder can be trained in parallel or reused, and the whole pipeline never needs end-to-end backpropagation through all layers. This is what enables the claimed reduction in training time and energy.

What would settle it

Train the two VAEs on data where the input field and solution have very different structure, such as a PDE with a sharp front that moves in time, keep the intermediate network fixed in size, and measure whether the latent-to-latent mapping error grows large enough that end-to-end training of the same components clearly outperforms the separable scheme. A smaller-scale check: swap the decoder VAE for one trained on a different but similar PDE and see whether the mapping still reconstructs solutions accurately.

Watch

Extended reading notes

Core claim

The central claim is that the three components of a surrogate model—an encoder that compresses the input field $y(\bm{x})$, a fully connected network that maps latent codes, and a decoder that reconstructs the solution $h(\bm{x},t)$—can be trained independently without sacrificing accuracy. Independence is achieved by training the encoder inside a VAE for $y(\bm{x})$ and the decoder inside a VAE for $h(\bm{x},t)$, so the intermediate network is the only piece that sees both latent spaces at once. The authors report that VAE-DNN outperforms FNO and DeepONet in both forward and inverse solutions of the nonlinear diffusion equation that governs groundwater flow in an unconfined aquifer, and does so with lower training time and energy use.

Load-bearing premise

The scheme works only if the two independently trained latent spaces land in the same coordinate system closely enough that a simple fully connected network can learn to translate between them; if they are misaligned, separable training cannot match an end-to-end model.

Editorial extensions

If this is right

  • Separable training cuts both the wall-clock time and the energy consumed during training relative to FNO and DeepONet.
  • The same separable scheme achieves lower error in forward predictions of the groundwater flow solution.
  • The same scheme achieves lower error in inverse estimates of the PDE parameters.
  • The encoder and decoder can be trained on their respective fields independently, so the intermediate mapping is the only component that must be trained on matched input–output pairs.

Reading between the lines

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

  • A natural extension the authors do not pursue is pretraining the two VAEs on unrelated datasets and fine-tuning only the fully connected bridge; whether the latent alignment survives such transfer is an open empirical question.
  • The energy comparison is specific to the chosen FNO and DeepONet implementations and problem sizes; the claimed advantage may shrink or grow on other architectures, PDEs, and hardware.
  • If real-world input and solution fields have very different dimensionality or structure, a learned linear or affine alignment layer between the two latent spaces might be needed to keep the separable training competitive.
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 / 2 minor

Summary. The abstract and title of the submitted manuscript, arXiv:2508.03839 (cs.LG), describe a trainable-by-parts surrogate model called VAE-DNN for solving forward and inverse parametric nonlinear PDEs, claiming superior accuracy and lower training time/energy compared to FNO and DeepONet on an unconfined aquifer groundwater flow problem. However, the supplied full text is a completely different paper, arXiv:2508.03833v3, titled "Computable Bounds for Strong Approximations with Applications" by Haoyu Ye and Morgane Austern, which develops a computable version of the Komlós–Major–Tusnády inequality. The submitted manuscript therefore contains an abstract that makes substantive machine-learning claims and a body that contains no description of the VAE-DNN architecture, no training procedure, no experimental results, and no comparison to FNO or DeepONet. The central claims of the paper as submitted cannot be checked because the promised content is absent and replaced by an unrelated probability-theory manuscript.

Significance. If the VAE-DNN claims were substantiated, the paper could be significant for the operator-learning community by offering an energy-efficient, modular alternative to FNO and DeepONet. The idea of separating the training of the encoder and decoder VAEs from the latent-space mapper is potentially interesting and could be a meaningful contribution if accompanied by a full specification and rigorous experiments. However, as submitted, the manuscript provides none of that content. The actual text is a probability paper with no connection to the abstract's claims; no architecture, proofs, code, or falsifiable predictions related to VAE-DNN are present. Therefore, the significance cannot be assessed beyond the abstract's assertions, which are unsupported by the submitted body.

major comments (4)
  1. [Abstract vs. Full Text] The abstract describes a VAE-DNN surrogate model for parametric PDEs, but the full text of the submission is an unrelated paper on computable KMT bounds (arXiv:2508.03833v3). This is an internal inconsistency that makes the central claim of the paper impossible to evaluate. The load-bearing elements named in the abstract—two VAEs, a fully connected latent-space mapper, comparisons to FNO and DeepONet, and training time/energy measurements—are entirely absent from the submitted body.
  2. [Full Text (Main Body)] No section of the submitted text contains the VAE-DNN architecture, the definition of the latent spaces $\bm\mu_{\phi_y}$ and $\bm\mu_{\phi_h}$, the training algorithm, or any specification of the encoder, decoder, or intermediate network. Without these components, the reader cannot verify the central methodological claim that the three parts can be trained independently on the basis of the two VAE latent spaces.
  3. [Full Text (Experiments)] The abstract reports findings that VAE-DNN delivers superior accuracy in both forward and inverse solutions for the nonlinear diffusion equation and demonstrates greater efficiency in time and energy. None of these results appear in the submitted manuscript. There are no tables, figures, error bars, or measurements of any kind related to the groundwater flow problem, so the performance claims are unsupported.
  4. [Full Text (Discussion)] The Discussion section of the supplied text concerns the asymptotic rate of the KMT bounds and contains the statement that the bound has a suboptimal logarithmic factor. This is a limitation of the KMT paper, not of the VAE-DNN model described in the abstract. The mismatch between the claimed subject and the actual content prevents any assessment of the VAE-DNN method's soundness or limitations.
minor comments (2)
  1. [Metadata] The arXiv identifier quoted in the submission materials (2508.03839) does not match the identifier shown in the full text (arXiv:2508.03833v3), which is a separate paper in mathematical statistics. The editor should verify the correct manuscript was uploaded.
  2. [Notation] The abstract uses notation such as $\bm\mu_{\phi_y}$ and $\bm\mu_{\phi_h}$, but these symbols are never defined or used anywhere in the supplied body, further indicating that the abstract and body originated from different works.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation is present because the submitted full text is an unrelated probability paper; the claimed VAE-DNN derivation and experiments are absent rather than circular.

full rationale

The abstract describes a trainable-by-parts VAE-DNN surrogate for parametric PDEs and claims accuracy and energy advantages over FNO and DeepONet. The full text, however, is a paper by Ye and Austern developing computable KMT coupling bounds; it contains no encoder, decoder, latent-space mapping, aquifer PDE experiments, or training-energy measurements. There is therefore no derivation chain in the submitted manuscript whose inputs can be shown equivalent to its outputs by construction. No fitted parameter is renamed as a prediction, no self-citation carries a load-bearing premise, and no ansatz is smuggled in via citation. The abstract-to-body mismatch is a severe completeness and consistency problem, but it is not a circularity: the claimed VAE-DNN derivation is missing, not self-referential. Under the requirement that circularity be demonstrated by quotable reduction, no such reduction exists here, so the honest finding is a score of 0.

Assumptions & free parameters 0 free parameters · 1 assumptions · 0 invented entities

No free parameters or invented entities can be identified from the abstract alone. One domain assumption about latent space compatibility is recorded because the separable training scheme relies on it.

assumptions (1)
  • domain assumption The latent spaces of the two VAEs are smooth and compatible enough that a fully connected network can map between them accurately.
    The method's success depends on the ability of a simple fully connected network to connect the input latent representation to the solution latent representation, which is not guaranteed for arbitrary PDE families.

how reviews work

0 comments
Cite this review

Pith. "Pith review of VAE-DNN: Energy-Efficient Trainable-by-Parts Surrogate Model For Parametric Partial Differential Equations." pith.science (2026). https://pith.science/paper/TOZJ4OPV

@misc{pith2026250803839,
  author       = {Pith},
  title        = {Pith review of: VAE-DNN: Energy-Efficient Trainable-by-Parts Surrogate Model For Parametric Partial Differential Equations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TOZJ4OPV}},
  note         = {Machine review of arXiv:2508.03839}
}
abstract

We propose a trainable-by-parts surrogate model for solving forward and inverse parameterized nonlinear partial differential equations. Like several other surrogate and operator learning models, the proposed approach employs an encoder to reduce the high-dimensional input $y(\bm{x})$ to a lower-dimensional latent space, $\bm\mu_{\bm\phi_y}$. Then, a fully connected neural network is used to map $\bm\mu_{\bm\phi_y}$ to the latent space, $\bm\mu_{\bm\phi_h}$, of the PDE solution $h(\bm{x},t)$. Finally, a decoder is utilized to reconstruct $h(\bm{x},t)$. The innovative aspect of our model is its ability to train its three components independently. This approach leads to a substantial decrease in both the time and energy required for training when compared to leading operator learning models such as FNO and DeepONet. The separable training is achieved by training the encoder as part of the variational autoencoder (VAE) for $y(\bm{x})$ and the decoder as part of the $h(\bm{x},t)$ VAE. We refer to this model as the VAE-DNN model. VAE-DNN is compared to the FNO and DeepONet models for obtaining forward and inverse solutions to the nonlinear diffusion equation governing groundwater flow in an unconfined aquifer. Our findings indicate that VAE-DNN not only demonstrates greater efficiency but also delivers superior accuracy in both forward and inverse solutions compared to the FNO and DeepONet models.

Discussion (0). Sign in to comment.

Forward citations

Cited by 2 Pith papers

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

  1. Latent Diffusion Posterior Sampling with Surrogate Likelihood Guidance for PDE Inverse Problems

    cs.CE 2026-06 unverdicted novelty 5.0 of 10

    L-DPS performs approximate Bayesian inference for PDE inverse problems by combining latent diffusion priors with surrogate-guided likelihood in a lower-dimensional space, demonstrated on Darcy flow permeability inversion.

  2. A Numerical Procedure for the Determination of the Pursuit Curve of Objects with Uniformly Accelerated Motion

    physics.class-ph 2025-08 unverdicted novelty 4.0 of 10

    The abstract claims equivalence results for accelerated pursuit curves and a numerical procedure for arbitrary paths, but the submission's full text is an unrelated machine-learning paper, so the claims cannot be checked.

Reference graph

Works this paper leans on

68 extracted references · 49 canonical work pages · cited by 2 Pith papers

  1. [1]

    Springer, 2003

    Søren Asmussen.Applied probability and queues. Springer, 2003

    work page 2003

  2. [2]

    Efficient concentration with Gaussian approximation.arXiv preprint arXiv:2208.09922, 2023

    Morgane Austern and Lester Mackey. Efficient concentration with Gaussian approximation.arXiv preprint arXiv:2208.09922, 2023

    arXiv 2023

  3. [3]

    Change-point analysis in financial networks.Stat, 9(1):e269, 2020

    Sayantan Banerjee and Kousik Guhathakurta. Change-point analysis in financial networks.Stat, 9(1):e269, 2020

    work page 2020

  4. [4]

    Koml´ os–Major–Tusn´ ady approximation under dependence.The Annals of Probability, 42(2):794–817, 2014

    Istv´ an Berkes, Weidong Liu, and Wei Biao Wu. Koml´ os–Major–Tusn´ ady approximation under dependence.The Annals of Probability, 42(2):794–817, 2014

    work page 2014

  5. [5]

    On strong embeddings by Stein’s method.Electronic Journal of Probability, 21(none):1 – 30, 2016

    Chinmoy Bhattacharjee and Larry Goldstein. On strong embeddings by Stein’s method.Electronic Journal of Probability, 21(none):1 – 30, 2016. 17

    work page 2016

  6. [6]

    Z. W. Birnbaum. An Inequality for Mill’s Ratio.The Annals of Mathematical Statistics, 13(2):245 – 246, 1942

    work page 1942

  7. [7]

    Rates in the Central Limit Theorem and diffusion approximation via Stein’s Method.arXiv preprint arXiv:1506.06966, 2015

    Thomas Bonis. Rates in the Central Limit Theorem and diffusion approximation via Stein’s Method.arXiv preprint arXiv:1506.06966, 2015

    arXiv 2015

  8. [8]

    Thomas Bonis. Stein’s method for normal approximation in wasserstein distances with application to the multivariate central limit theorem.Probability Theory and Related Fields, 178(3):827–860, 2020

    2020

Show all 68 references
  1. [9]

    Improved rates of convergence for the multivariate central limit theorem in wasser- stein distance.arXiv preprint arXiv:2305.14248, 2023

    Thomas Bonis. Improved rates of convergence for the multivariate central limit theorem in wasser- stein distance.arXiv preprint arXiv:2305.14248, 2023

    2023 arXiv

  2. [10]

    Automatic change detection in multimodal serial MRI: application to multiple sclerosis lesion evolution.NeuroImage, 20(2):643–656, 2003

    Marcel Bosc, Fabrice Heitz, Jean-Paul Armspach, Izzie Namer, Daniel Gounot, and Lucien Rum- bach. Automatic change detection in multimodal serial MRI: application to multiple sclerosis lesion evolution.NeuroImage, 20(2):643–656, 2003

    2003

  3. [11]

    Simulation of Brownian motion at first-passage times.Math- ematics and Computers in Simulation, 77(1):64–71, 2008

    Zaeem A Burq and Owen D Jones. Simulation of Brownian motion at first-passage times.Math- ematics and Computers in Simulation, 77(1):64–71, 2008

    2008

  4. [12]

    Bounds on the running maximum of a random walk with small drift.arXiv preprint arXiv:2010.08767, 2020

    Ofer Busani and Timo Sepp¨ al¨ ainen. Bounds on the running maximum of a random walk with small drift.arXiv preprint arXiv:2010.08767, 2020

    2010 arXiv

  5. [13]

    Strong approximations of bivariate uniform empirical processes

    Nathalie Castelle and Fran¸ coise Laurent-Bonvalot. Strong approximations of bivariate uniform empirical processes. InAnnales de l’Institut Henri Poincare (B) Probability and Statistics, vol- ume 34, pages 425–480. Elsevier, 1998

    1998

  6. [14]

    A new approach to strong embeddings.Probability Theory and Related Fields, 152(1-2):231–264, 2012

    Sourav Chatterjee. A new approach to strong embeddings.Probability Theory and Related Fields, 152(1-2):231–264, 2012

    2012

  7. [15]

    Inference of breakpoints in high-dimensional time series.Journal of the American Statistical Association, 117(540):1951–1963, 2022

    Likai Chen, Weining Wang, and Wei Biao Wu. Inference of breakpoints in high-dimensional time series.Journal of the American Statistical Association, 117(540):1951–1963, 2022

    1951

  8. [16]

    From stein identities to moderate deviations

    Louis HY Chen, Xiao Fang, and Qi-Man Shao. From stein identities to moderate deviations. 2013

    2013

  9. [17]

    Some applications of first-passage ideas to finance, 2013

    R´ emy Chicheportiche and Jean-Philippe Bouchaud. Some applications of first-passage ideas to finance, 2013

    2013

  10. [18]

    Sharp bounds on the absolute moments of a sum of two iid random variables.The Annals of Probability, pages 765–771, 1983

    David C Cox and Johannes HB Kemperman. Sharp bounds on the absolute moments of a sum of two iid random variables.The Annals of Probability, pages 765–771, 1983

    1983

  11. [19]

    Sur un nouveau th´ eoreme-limite de la th´ eorie des probabilit´ es

    Harald Cram´ er. Sur un nouveau th´ eoreme-limite de la th´ eorie des probabilit´ es. InCollected Works II, pages 895–913. Springer, 1994

    1994

  12. [20]

    Simulation of first-passage times for alternating Brownian motions.Methodology and Computing in Applied Probability, 7(2):161–181, 2005

    Antonio Di Crescenzo, Elvira Di Nardo, and LM Ricciardi. Simulation of first-passage times for alternating Brownian motions.Methodology and Computing in Applied Probability, 7(2):161–181, 2005

    2005

  13. [21]

    Bounds for the absolute third moment.Scandinavian Journal of Statistics, pages 149–152, 1975

    Carl-Gustav Esseen. Bounds for the absolute third moment.Scandinavian Journal of Statistics, pages 149–152, 1975

    1975

  14. [22]

    Fromp-wasserstein bounds to moderate deviations.Electronic Journal of Probability, 28:1–52, 2023

    Xiao Fang and Yuta Koike. Fromp-wasserstein bounds to moderate deviations.Electronic Journal of Probability, 28:1–52, 2023

    2023

  15. [23]

    Retracted chapter: Generalization of a probability limit theorem of cram´ er

    William Feller. Retracted chapter: Generalization of a probability limit theorem of cram´ er. In Selected Papers I, pages 601–612. Springer, 2015

    2015

  16. [24]

    Multiscale change point inference.Journal of the Royal Statistical Society Series B: Statistical Methodology, 76(3):495–580, 2014

    Klaus Frick, Axel Munk, and Hannes Sieling. Multiscale change point inference.Journal of the Royal Statistical Society Series B: Statistical Methodology, 76(3):495–580, 2014

    2014

  17. [25]

    Random walk models for the spike activity of a single neuron.Biophysical journal, 4(1):41–68, 1964

    George L Gerstein and Benoit Mandelbrot. Random walk models for the spike activity of a single neuron.Biophysical journal, 4(1):41–68, 1964. 18

    1964

  18. [26]

    Cambridge University Press, 2014

    Wulfram Gerstner, Werner M Kistler, Richard Naud, and Liam Paninski.Neuronal dynamics: From single neurons to networks and models of cognition. Cambridge University Press, 2014

    2014

  19. [27]

    Bayesian online change point detection in finance.Financial Internet Quarterly, 17(4):27–33, 2021

    Reza Habibi. Bayesian online change point detection in finance.Financial Internet Quarterly, 17(4):27–33, 2021

    2021

  20. [28]

    Springer Science & Business Media, 2012

    Douglas M Hawkins and David H Olwell.Cumulative sum charts and charting for quality im- provement. Springer Science & Business Media, 2012

    2012

  21. [29]

    Courier Corporation, 2012

    Ronald A Howard.Dynamic Probabilistic Systems, Volume I: Markov Models, volume 1. Courier Corporation, 2012

    2012

  22. [30]

    Self-normalized cram´ er-type large deviations for independent random variables.The Annals of probability, 31(4):2167–2215, 2003

    Bing-Yi Jing, Qi-Man Shao, and Qiying Wang. Self-normalized cram´ er-type large deviations for independent random variables.The Annals of probability, 31(4):2167–2215, 2003

    2003

  23. [31]

    Springer Science & Business Media, 2012

    Ioannis Karatzas and Steven Shreve.Brownian motion and stochastic calculus, volume 113. Springer Science & Business Media, 2012

    2012

  24. [32]

    Dynkin ′ s Games and Israeli Options.International Scholarly Research Notices, 2013(1):856458, 2013

    Yuri Kifer. Dynkin ′ s Games and Israeli Options.International Scholarly Research Notices, 2013(1):856458, 2013

    2013

  25. [33]

    An approximation of partial sums of independent R V’-s, and the sample DF

    J´ anos Koml´ os, P´ eter Major, and G´ abor Tusn´ ady. An approximation of partial sums of independent R V’-s, and the sample DF. I.Zeitschrift f¨ ur Wahrscheinlichkeitstheorie und verwandte Gebiete, 32(1):111–131, 1975

    1975

  26. [34]

    An approximation of partial sums of independent R V’s, and the sample DF

    J´ anos Koml´ os, P´ eter Major, and G´ abor Tusn´ ady. An approximation of partial sums of independent R V’s, and the sample DF. II.Zeitschrift f¨ ur Wahrscheinlichkeitstheorie und verwandte Gebiete, 34(1):33–58, 1976

    1976

  27. [35]

    A jump-diffusion model for option pricing.Management science, 48(8):1086–1101, 2002

    Steven G Kou. A jump-diffusion model for option pricing.Management science, 48(8):1086–1101, 2002

    2002

  28. [36]

    First exit time of a random walk from the boundsf(n)±cg(n), with applications.The Annals of Probability, 7(4):672–692, 1979

    TL Lai and RA Wijsman. First exit time of a random walk from the boundsf(n)±cg(n), with applications.The Annals of Probability, 7(4):672–692, 1979

    1979

  29. [37]

    Optimal stopping and embedding.Journal of applied probability, 37(4):1143–1148, 2000

    Damien Lamberton and LCG Rogers. Optimal stopping and embedding.Journal of applied probability, 37(4):1143–1148, 2000

    2000

  30. [38]

    Habilitation ` a diriger des recherches, Universit´ e de Lille Nord de France, February 2019

    Odalric-Ambrym Maillard.Mathematics of Statistical Sequential Decision Making. Habilitation ` a diriger des recherches, Universit´ e de Lille Nord de France, February 2019

    2019

  31. [39]

    The approximation of partial sums of independent rv’s.Zeitschrift f¨ ur Wahrschein- lichkeitstheorie und verwandte Gebiete, 35(3):213–220, 1976

    P´ eter Major. The approximation of partial sums of independent rv’s.Zeitschrift f¨ ur Wahrschein- lichkeitstheorie und verwandte Gebiete, 35(3):213–220, 1976

    1976

  32. [40]

    Online Bayesian change point detection algorithms for segmentation of epileptic activity

    Rakesh Malladi, Giridhar P Kalamangalam, and Behnaam Aazhang. Online Bayesian change point detection algorithms for segmentation of epileptic activity. In2013 Asilomar conference on signals, systems and computers, pages 1833–1837. IEEE, 2013

    2013

  33. [41]

    Sharp empirical Bernstein bounds for the variance of bounded random variables.arXiv preprint arXiv:2505.01987, 2025

    Diego Martinez-Taboada and Aaditya Ramdas. Sharp empirical Bernstein bounds for the variance of bounded random variables.arXiv preprint arXiv:2505.01987, 2025

    2025 arXiv

  34. [42]

    Quantile coupling inequalities and their applications

    David M Mason and Harrison H Zhou. Quantile coupling inequalities and their applications. 2012

    2012

  35. [43]

    Sequential Gaussian approximation for nonstationary time series in high dimensions.Bernoulli, 29(4):3114–3140, 2023

    Fabian Mies and Ansgar Steland. Sequential Gaussian approximation for nonstationary time series in high dimensions.Bernoulli, 29(4):3114–3140, 2023

    2023

  36. [44]

    John wiley & sons, 2020

    Douglas C Montgomery.Introduction to statistical quality control. John wiley & sons, 2020

    2020

  37. [45]

    Some inequalities for the distribution of sums of independent random variables.Theory of Probability & Its Applications, 22(2):248–256, 1978

    SV Nagaev and IF Pinelis. Some inequalities for the distribution of sums of independent random variables.Theory of Probability & Its Applications, 22(2):248–256, 1978

    1978

  38. [46]

    Asymptotic equivalence of density estimation and Gaussian white noise.The Annals of Statistics, pages 2399–2430, 1996

    Michael Nussbaum. Asymptotic equivalence of density estimation and Gaussian white noise.The Annals of Statistics, pages 2399–2430, 1996. 19

    1996

  39. [47]

    A note on Burkholder-Rosenthal inequality.Bull

    Adam Osekowski. A note on Burkholder-Rosenthal inequality.Bull. Pol. Acad. Sci. Math, 60(2):177–185, 2012

    2012

  40. [48]

    Generalization of an inequality by Talagrand and links with the logarithmic Sobolev inequality.Journal of Functional Analysis, 173(2):361–400, 2000

    Felix Otto and C´ edric Villani. Generalization of an inequality by Talagrand and links with the logarithmic Sobolev inequality.Journal of Functional Analysis, 173(2):361–400, 2000

    2000

  41. [49]

    Continuous inspection schemes.Biometrika, 41(1/2):100–115, 1954

    Ewan S Page. Continuous inspection schemes.Biometrika, 41(1/2):100–115, 1954

    1954

  42. [50]

    V. V. Petrov.Sums of Independent Random Variables. De Gruyter, Berlin, Boston, 1975

    1975

  43. [51]

    Springer Science & Business Media, 2012

    Valentin V Petrov.Sums of independent random variables, volume 82. Springer Science & Business Media, 2012

    2012

  44. [52]

    Probability bounds for first exits through moving boundaries.The Annals of Probability, pages 106–117, 1978

    Stephen Portnoy. Probability bounds for first exits through moving boundaries.The Annals of Probability, pages 106–117, 1978

    1978

  45. [53]

    John Wiley & Sons, 2003

    Marvin Rausand and Arnljot Hoyland.System reliability theory: models, statistical methods, and applications, volume 396. John Wiley & Sons, 2003

    2003

  46. [54]

    Moment inequalities for sums of dependent random variables under projective conditions.Journal of Theoretical Probability, 22(1):146–163, 2009

    Emmanuel Rio. Moment inequalities for sums of dependent random variables under projective conditions.Journal of Theoretical Probability, 22(1):146–163, 2009

    2009

  47. [55]

    A remark on Stirling’s formula.The American mathematical monthly, 62(1):26– 29, 1955

    Herbert Robbins. A remark on Stirling’s formula.The American mathematical monthly, 62(1):26– 29, 1955

    1955

  48. [56]

    On the accuracy of normal approximation in the invariance principle.Matematicheskie Trudy, 13:40–66, 1989

    Aleksandr Ivanovich Sakhanenko. On the accuracy of normal approximation in the invariance principle.Matematicheskie Trudy, 13:40–66, 1989

    1989

  49. [57]

    Reducing sequential change detection to sequential estimation.arXiv preprint arXiv:2309.09111, 2023

    Shubhanshu Shekhar and Aaditya Ramdas. Reducing sequential change detection to sequential estimation.arXiv preprint arXiv:2309.09111, 2023

    2023 arXiv

  50. [58]

    Springer, 2004

    Steven E Shreve et al.Stochastic calculus for finance II: Continuous-time models, volume 11. Springer, 2004

    2004

  51. [59]

    Staudacher, S

    M. Staudacher, S. Telser, A. Amann, H. Hinterhuber, and M. Ritsch-Marte. A new method for change-point detection developed for on-line analysis of the heart beat variability during sleep. Physica A: Statistical Mechanics and its Applications, 349(3):582–596, 2005

    2005

  52. [60]

    John Wiley & Sons, 2001

    Kishor S Trivedi.Probability and statistics with reliability, queuing, and computer science appli- cations. John Wiley & Sons, 2001

    2001

  53. [61]

    Survival Multiarmed Bandits with Bootstrapping Methods

    Peter Veroutis and Fr´ ed´ eric Godin. Survival Multiarmed Bandits with Bootstrapping Methods. arXiv preprint arXiv:2410.16486, 2024

    2024 arXiv

  54. [62]

    American Mathematical Soc., 2003

    C´ edric Villani.Topics in Optimal Transportation, volume 58. American Mathematical Soc., 2003

    2003

  55. [63]

    The rate of convergence of the binomial tree scheme.Finance and Stochastics, 7(3):337–361, 2003

    John B Walsh. The rate of convergence of the binomial tree scheme.Finance and Stochastics, 7(3):337–361, 2003

    2003

  56. [64]

    Distribution-uniform anytime- valid sequential inference.arXiv preprint arXiv:2311.03343, 2023

    Ian Waudby-Smith, Edward H Kennedy, and Aaditya Ramdas. Distribution-uniform anytime- valid sequential inference.arXiv preprint arXiv:2311.03343, 2023

    2023

  57. [65]

    Nonasymptotic and distribution- uniform Koml´ os–Major–Tusn´ ady approximation.arXiv preprint arXiv:2502.06188, 2025

    Ian Waudby-Smith, Martin Larsson, and Aaditya Ramdas. Nonasymptotic and distribution- uniform Koml´ os–Major–Tusn´ ady approximation.arXiv preprint arXiv:2502.06188, 2025

    2025

  58. [66]

    Estimating means of bounded random variables by betting.Journal of the Royal Statistical Society Series B: Statistical Methodology, 86(1):1–27, 2024

    Ian Waudby-Smith and Aaditya Ramdas. Estimating means of bounded random variables by betting.Journal of the Royal Statistical Society Series B: Statistical Methodology, 86(1):1–27, 2024

    2024

  59. [67]

    Some current directions in the theory and application of statistical process monitoring.Journal of quality technology, 46(1):78–94, 2014

    William H Woodall and Douglas C Montgomery. Some current directions in the theory and application of statistical process monitoring.Journal of quality technology, 46(1):78–94, 2014

    2014

  60. [68]

    Adaptive change detection in heart rate trend monitoring in anesthetized children.IEEE transactions on biomedical engineering, 53(11):2211–2219, 2006

    Ping Yang, Guy Dumont, and John Mark Ansermino. Adaptive change detection in heart rate trend monitoring in anesthetized children.IEEE transactions on biomedical engineering, 53(11):2211–2219, 2006. 20

    2006

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.