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
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 →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
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.
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
- 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.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
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)
- [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.
- [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.
- [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.
- [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)
- [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.
- [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
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
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.
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.
Forward citations
Cited by 2 Pith papers
-
Latent Diffusion Posterior Sampling with Surrogate Likelihood Guidance for PDE Inverse Problems
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.
-
A Numerical Procedure for the Determination of the Pursuit Curve of Objects with Uniformly Accelerated Motion
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
- [1]
-
[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]
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]
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]
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]
Z. W. Birnbaum. An Inequality for Mill’s Ratio.The Annals of Mathematical Statistics, 13(2):245 – 246, 1942
work page 1942
-
[7]
Thomas Bonis. Rates in the Central Limit Theorem and diffusion approximation via Stein’s Method.arXiv preprint arXiv:1506.06966, 2015
arXiv 2015
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[16]
From stein identities to moderate deviations
Louis HY Chen, Xiao Fang, and Qi-Man Shao. From stein identities to moderate deviations. 2013
2013
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[29]
Courier Corporation, 2012
Ronald A Howard.Dynamic Probabilistic Systems, Volume I: Markov Models, volume 1. Courier Corporation, 2012
2012
-
[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
-
[31]
Springer Science & Business Media, 2012
Ioannis Karatzas and Steven Shreve.Brownian motion and stochastic calculus, volume 113. Springer Science & Business Media, 2012
2012
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[42]
Quantile coupling inequalities and their applications
David M Mason and Harrison H Zhou. Quantile coupling inequalities and their applications. 2012
2012
-
[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
-
[44]
John wiley & sons, 2020
Douglas C Montgomery.Introduction to statistical quality control. John wiley & sons, 2020
2020
-
[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
-
[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
-
[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
-
[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
-
[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
-
[50]
V. V. Petrov.Sums of Independent Random Variables. De Gruyter, Berlin, Boston, 1975
1975
-
[51]
Springer Science & Business Media, 2012
Valentin V Petrov.Sums of independent random variables, volume 82. Springer Science & Business Media, 2012
2012
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
-
[58]
Springer, 2004
Steven E Shreve et al.Stochastic calculus for finance II: Continuous-time models, volume 11. Springer, 2004
2004
-
[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
-
[60]
John Wiley & Sons, 2001
Kishor S Trivedi.Probability and statistics with reliability, queuing, and computer science appli- cations. John Wiley & Sons, 2001
2001
-
[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
-
[62]
American Mathematical Soc., 2003
C´ edric Villani.Topics in Optimal Transportation, volume 58. American Mathematical Soc., 2003
2003
-
[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
-
[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
-
[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
-
[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
-
[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
-
[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
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.