Pith. sign in

REVIEW 3 major objections 3 minor 58 references

Robust Simulation Based Inference

T0 review · 3 major / 3 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 Simulation-based inference can produce valid confidence sets even when the model is misspecified and regularity conditions fail.

desk verdict Robust SBI paper with a plausible framework and strong coverage claims, but the full text is unreadable so the central guarantees are unverifiable from this submission. read the letter →

arxiv 2508.02404 v1 pith:NDHZIQO6 submitted 2025-08-04 stat.ME

Lorenzo Tomaselli , Valérie Ventura , Larry Wasserman This is my paper
classification stat.ME MSC 62F3562F2562G10
keywords simulation-basedinferencemodelmisspecificationprojectionparameterfrequentistcoverageexponentialtiltinggoodness-of-fittestintractablelikelihoodactivelearning
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

Simulation-based inference (SBI) usually assumes the simulator generates data exactly like the real process; when that assumption is wrong, the resulting confidence sets can badly undercover. This paper claims a way to keep valid frequentist guarantees under misspecification: instead of targeting the 'true' parameter, target the parameter that minimizes a chosen discrepancy between the real distribution and the model family, and build confidence sets for that projection parameter using simulations. The authors also offer exponential tilting to expand the model, a goodness-of-fit test to detect misspecification, and two tools to make SBI practical. A sympathetic reader should care because SBI is used precisely when likelihoods are intractable, and in those settings the model is usually a known approximation, not the truth.

What carries the argument

The central object is the projection parameter, the minimizer of a user-chosen discrepancy between the true distribution and the assumed model family. Because the target is defined by minimization rather than by the unknown truth, inference about it is well-posed under misspecification; the confidence sets are constructed directly from simulated replicas of this discrepancy, which is what supplies non-asymptotic validity without regularity conditions. The secondary mechanisms are exponential tilting to enlarge the model and a discrepancy-based goodness-of-fit test.

What would settle it

Simulate data from a distribution whose discrepancy to the assumed model family has two distinct global minima. If the proposed confidence set does not contain both minima with the nominal frequency, or if the method must arbitrarily select one minimizer for the guarantee to hold, the universal coverage claim fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that valid frequentist inference is possible in simulation-based inference even when the assumed model is wrong and standard regularity conditions fail. The target is redefined as the projection parameter $\theta^*$ that minimizes a discrepancy $D(P, P_\theta)$ between the true distribution $P$ and the assumed model family $\{P_\theta\}$. The method constructs confidence sets for $\theta^*$ whose coverage is guaranteed by the simulation mechanism itself, not by asymptotic theory. The paper further claims that exponential-tilting model expansion gives an alternative route to the same end, and that an SBI-based goodness-of-fit test can flag misspecification. If true, this removes the central limitation that has kept SBI confined to settings where the simulator is trusted as correct.

Load-bearing premise

The target of inference must be a unique, stable minimizer of the discrepancy; if multiple parameter values tie for the best fit, the claimed coverage guarantee has no single parameter to cover.

Editorial extensions

If this is right

  • Confidence sets for the best-fitting model parameter remain valid when the simulator is only an approximation, so SBI can be applied to real data without pretending the model is exact.
  • The guarantees do not depend on asymptotic normality or smoothness, so the method covers non-regular problems such as boundary parameters, singular models, or heavy-tailed data.
  • The goodness-of-fit test lets practitioners check whether their simulator is adequate before drawing conclusions from the confidence sets.
  • Model expansion by exponential tilting gives a data-driven way to improve a deficient model family while staying inside the simulation framework.
  • The proposed closed-form approximations and active learning sampling scheme make robust SBI computationally practical.

Reading between the lines

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

  • The projection-parameter viewpoint generalizes the classical 'pseudo-true' value of maximum likelihood under misspecification; choosing different discrepancies corresponds to different robustness/efficiency trade-offs, a connection the paper does not develop.
  • The goodness-of-fit test could double as a stopping rule for the active learning sampler: keep sampling parameters until the discrepancy test stops rejecting, closing the loop between the two ideas.
  • If the discrepancy is chosen to be an integral probability metric, the framework may connect to distributionally robust optimization, where the projection is exactly the robust decision under an ambiguity set.
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

3 major / 3 minor

Summary. The paper proposes a framework for robust simulation-based inference (SBI) under model misspecification. The target of inference is a projection parameter defined as the minimizer of a discrepancy between the true distribution and the assumed model. The abstract claims that the method guarantees valid frequentist inference even when the model is incorrectly specified and even when standard regularity conditions fail. The paper also introduces model expansion via exponential tilting, an SBI-based goodness-of-fit test, an approach to closed-form approximation of intractable models, and an active learning strategy for parameter-space sampling. The provided full text, however, is almost entirely undecodable mojibake, so the derivations, algorithm definitions, theorems, and any simulation results cannot be inspected. The assessment below is therefore based on the abstract and the visible fragments of the manuscript.

Significance. If the central guarantee is valid, the paper addresses an important and timely problem: standard SBI methods rely on the model being correctly specified, and misspecification can invalidate inference. A principled projection-based framework with finite-sample or asymptotic coverage guarantees under misspecification would be a substantial contribution to the SBI literature. The additional contributions—exponential tilting, a misspecification test, closed-form approximation, and active learning—could also be valuable if they are developed rigorously. However, the manuscript as submitted provides no readable technical content to verify these contributions. There are no visible derivations, proofs, or simulation results, and no machine-checked artifacts (such as code or formal proofs) are available in the text. The significance of the work therefore cannot be confirmed on the basis of this submission.

major comments (3)
  1. [Full Text] The entire body of the manuscript is presented as undecodable mojibake (for example, the opening line is '��������� �������������� �� ��������������������'). As a result, none of the mathematical definitions, algorithm descriptions, theorems, proofs, or simulation studies can be checked. This is a load-bearing problem for the central claim of the paper: the abstract asserts strong guarantees, but the evidence supporting those guarantees is entirely inaccessible. The manuscript cannot be evaluated in its current form, and the authors should be asked to provide a readable version before any substantive review can occur.
  2. [Abstract, final paragraph] The abstract claims that the method 'guarantees valid inference' under model misspecification and without standard regularity conditions, but it does not state any condition ensuring that the projection parameter is well-defined. In particular, the abstract does not assert uniqueness or identifiability of the minimizer of the discrepancy between the true distribution and the assumed model. If multiple parameters achieve the same minimal discrepancy, the target of inference is not a well-defined parameter, and any coverage statement about 'the' projection parameter becomes vacuous or dependent on arbitrary selection rules. This is precisely the kind of condition that the full text would need to state explicitly; the corrupted full text does not permit verification.
  3. [Abstract, coverage guarantee] Even if the projection parameter is uniquely defined, a valid frequentist coverage guarantee requires control of the estimation error of the minimizer under the true distribution. The abstract explicitly disclaims standard regularity conditions, but it provides neither finite-sample bounds nor a clearly specified asymptotic framework. Without such control, the claim that the method 'guarantees valid inference' is unsupported as stated. The full text, if readable, might supply these conditions, but the submitted manuscript does not.
minor comments (3)
  1. [Title] The title 'Robust Simulation Based Inference' would benefit from a hyphen: 'Robust Simulation-Based Inference'.
  2. [Abstract, first paragraph] The phrase 'as always, this can lead to invalid inference when the model is misspecified' is informal for a methods paper; consider rephrasing to something like 'misspecification can lead to invalid inference, as is well known.'
  3. [Full Text, bottom] The plain-text line 'arXiv:2508.02405v1 [cs.RO] 4 Aug 2025' appears at the end of the document and should be removed from the body in a final submission.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity identified in the readable abstract; the full text is mojibake and no derivation chain can be examined.

full rationale

The only readable portion of the manuscript is the abstract, which defines the target of inference as a projection parameter that minimizes a discrepancy between the true distribution and the assumed model. This is an externally defined quantity, not one defined through the method's own outputs, so the central confidence-set guarantee is not circular by construction. The remaining text is entirely mojibake, so no equations, proofs, or self-citations can be inspected to test for hidden circularity. The abstract's coverage claim may depend on unstated identifiability conditions for the discrepancy minimizer, but that is a correctness or assumptions concern, not evidence of circularity. No specific reduction of a prediction to a fitted input or self-citation chain can be quoted, so the appropriate finding is no significant circularity.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper's validity claim rests on the existence and identifiability of the projection parameter, and on the ability to simulate from the assumed model. These are not free parameters but structural assumptions required for the method to work.

assumptions (3)
  • domain assumption The assumed model can generate simulations and the true distribution is fixed.
    SBI requires the ability to sample from the assumed model; the abstract states this implicitly.
  • ad hoc to paper There exists a unique minimizer of the discrepancy between the true distribution and the assumed model over the parameter space.
    The central inference target is defined as this minimizer; validity depends on existence and identifiability, which is not assured in general.
  • domain assumption The frequentist validity guarantee holds for the chosen discrepancy and simulation error is asymptotically negligible or controlled.
    Guarantee of valid confidence sets relies on convergence of simulation-based estimates to population quantities.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust Simulation Based Inference." pith.science (2026). https://pith.science/paper/NDHZIQO6

@misc{pith2026250802404,
  author       = {Pith},
  title        = {Pith review of: Robust Simulation Based Inference},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NDHZIQO6}},
  note         = {Machine review of arXiv:2508.02404}
}
read the original abstract

Simulation-Based Inference (SBI) is an approach to statistical inference where simulations from an assumed model are used to construct estimators and confidence sets. SBI is often used when the likelihood is intractable and to construct confidence sets that do not rely on asymptotic methods or regularity conditions. Traditional SBI methods assume that the model is correct, but, as always, this can lead to invalid inference when the model is misspecified. This paper introduces robust methods that allow for valid frequentist inference in the presence of model misspecification. We propose a framework where the target of inference is a projection parameter that minimizes a discrepancy between the true distribution and the assumed model. The method guarantees valid inference, even when the model is incorrectly specified and even if the standard regularity conditions fail. Alternatively, we introduce model expansion through exponential tilting as another way to account for model misspecification. We also develop an SBI based goodness-of-fit test to detect model misspecification. Finally, we propose two ideas that are useful in the SBI framework beyond robust inference: an SBI based method to obtain closed form approximations of intractable models and an active learning approach to more efficiently sample the parameter space.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

58 extracted references · 48 canonical work pages

  1. [1]

    write newline

    " write newline " cite write " FUNCTION editor.postfix editor num.names #1 > "( )" "( )" if FUNCTION editor.trans.postfix editor num.names #1 > "( )" "( )" if FUNCTION trans.postfix translator num.names #1 > "( )" "( )" if FUNCTION authors.editors.reflist.apa5 'field := 'dot := field num.names 'numnames := numnames 'format.num.names := format.num.names na...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  3. [3]

    Robust and efficient estimation by minimising a density power divergence

    Basu, A., Harris, I.R., Hjort, N.L., and Jones, M.C. Robust and efficient estimation by minimising a density power divergence. Biometrika, 85 0 (3): 0 549--559, September 1998

    work page 1998

  4. [4]

    and Kohler, M

    Bauer, B. and Kohler, M. On deep learning as a remedy for the curse of dimensionality in nonparametric regression . The Annals of Statistics, 47 0 (4): 0 2261 -- 2285, 2019

    work page 2019

  5. [5]

    Approximate Bayesian Computation in Evolution and Ecology

    Beaumont, M.A. Approximate Bayesian Computation in Evolution and Ecology . Annual Review of Ecology, Evolution, and Systematics, 41 0 (Volume 41, 2010): 0 379--406, December 2010

    work page 2010

  6. [6]

    Approximate Bayesian Computation in Population Genetics

    Beaumont, M.A., Zhang, W., and Balding, D.J. Approximate Bayesian Computation in Population Genetics . Genetics, 162 0 (4): 0 2025--2035, December 2002

    work page 2025

  7. [7]

    Minimum Hellinger Distance Estimates for Parametric Models

    Beran, R. Minimum Hellinger Distance Estimates for Parametric Models . The Annals of Statistics, 5 0 (3): 0 445--463, 1977. Publisher: Institute of Mathematical Statistics

    work page 1977

  8. [8]

    and Ventura, L

    Bortolato, E. and Ventura, L. Box confidence depth: simulation-based inference with hyper-rectangles. arXiv preprint arXiv:2502.11072, 2025

    work page arXiv 2025

Show all 58 references
  1. [9]

    Mining gold from implicit models to improve likelihood-free inference

    Brehmer, J., Louppe, G., Pavez, J., and Cranmer, K. Mining gold from implicit models to improve likelihood-free inference. Proceedings of the National Academy of Sciences, 117 0 (10): 0 5242--5249, March 2020. Publisher: Proceedings of the National Academy of Sciences

    2020

  2. [10]

    Time Series Analysis via Mechanistic Models

    Bretó, C., He, D., Ionides, E.L., and King, A.A. Time Series Analysis via Mechanistic Models . The Annals of Applied Statistics, 3 0 (1): 0 319--348, 2009. Publisher: Institute of Mathematical Statistics

    2009

  3. [11]

    Statistical Inference for Generative Models with Maximum Mean Discrepancy , June 2019

    Briol, F.X., Barp, A., Duncan, A.B., and Girolami, M. Statistical Inference for Generative Models with Maximum Mean Discrepancy , June 2019. arXiv:1906.05944 [cs, math, stat]

    2019 arXiv

  4. [12]

    lhs, June 2023

    Carnell, R. lhs, June 2023. original-date: 2018-12-01T20:59:29Z

    2023

  5. [13]

    and Alquier, P

    Ch \'e rief-Abdellatif, B.E. and Alquier, P. Finite sample properties of parametric mmd estimation: robustness to misspecification and dependence. Bernoulli, 28 0 (1): 0 181--213, 2022

    2022

  6. [14]

    Statistical optimal transport

    Chewi, S., Niles-Weed, J., and Rigollet, P. Statistical optimal transport. arXiv preprint arXiv:2407.18163, 2024

    2024 arXiv

  7. [15]

    Approximating Likelihood Ratios with Calibrated Discriminative Classifiers , March 2016

    Cranmer, K., Pavez, J., and Louppe, G. Approximating Likelihood Ratios with Calibrated Discriminative Classifiers , March 2016. arXiv:1506.02169 [physics, stat]

    2016 arXiv

  8. [16]

    The frontier of simulation-based inference

    Cranmer, K., Brehmer, J., and Louppe, G. The frontier of simulation-based inference. Proceedings of the National Academy of Sciences, 117 0 (48): 0 30055--30062, December 2020. Publisher: Proceedings of the National Academy of Sciences

    2020

  9. [17]

    Likelihood- Free Frequentist Inference : Confidence Sets with Correct Conditional Coverage , April 2023

    Dalmasso, N., Masserano, L., Zhao, D., Izbicki, R., and Lee, A.B. Likelihood- Free Frequentist Inference : Confidence Sets with Correct Conditional Coverage , April 2023. arXiv:2107.03920 [cs, stat]

    2023 arXiv

  10. [18]

    and Guillin, A

    Fournier, N. and Guillin, A. On the rate of convergence in wasserstein distance of the empirical measure. Probability theory and related fields, 162 0 (3): 0 707--738, 2015

    2015

  11. [19]

    Large sample analysis of the median heuristic

    Garreau, D., Jitkrittum, W., and Kanagawa, M. Large sample analysis of the median heuristic. arXiv preprint arXiv:1707.07269, 2017

    2017 arXiv

  12. [20]

    and Ramdas, A

    Gasparin, M. and Ramdas, A. Merging uncertainty sets via majority vote, March 2024. arXiv:2401.09379

    2024 arXiv

  13. [21]

    Accelerating Bayesian inference for stochastic epidemic models using incidence data, March 2023

    Golightly, A., Wadkin, L.E., Whitaker, S., Baggaley, A., Parker, N., and Kypraios, T. Accelerating Bayesian inference for stochastic epidemic models using incidence data, March 2023. arXiv:2303.15371 [stat]

    2023 arXiv

  14. [22]

    A Kernel Two - Sample Test

    Gretton, A., Borgwardt, K.M., Rasch, M.J., Schölkopf, B., and Smola, A. A Kernel Two - Sample Test . Journal of Machine Learning Research, 13 0 (25): 0 723--773, 2012

    2012

  15. [23]

    Multivariate goodness-of-fit tests based on wasserstein distance

    Hallin, M., Mordant, G., and Segers, J. Multivariate goodness-of-fit tests based on wasserstein distance. 2021

    2021

  16. [24]

    Reconstruction of the full transmission dynamics of COVID -19 in Wuhan

    Hao, X., Cheng, S., Wu, D., Wu, T., Lin, X., and Wang, C. Reconstruction of the full transmission dynamics of COVID -19 in Wuhan . Nature, 584 0 (7821): 0 420--424, August 2020. Number: 7821 Publisher: Nature Publishing Group

    2020

  17. [25]

    Inference for nonlinear dynamical systems

    Ionides, E.L., Bretó, C., and King, A.A. Inference for nonlinear dynamical systems. Proceedings of the National Academy of Sciences, 103 0 (49): 0 18438--18443, December 2006. Publisher: Proceedings of the National Academy of Sciences

    2006

  18. [26]

    Inference for dynamic and latent variable models via iterated, perturbed Bayes maps

    Ionides, E.L., Nguyen, D., Atchadé, Y., Stoev, S., and King, A.A. Inference for dynamic and latent variable models via iterated, perturbed Bayes maps. Proceedings of the National Academy of Sciences, 112 0 (3): 0 719--724, January 2015

    2015

  19. [27]

    and Turnbull, B

    Jiang, W. and Turnbull, B. The Indirect Method : Inference Based on Intermediate Statistics — A Synthesis and Examples . Statistical Science, 19 0 (2): 0 239--263, May 2004. Publisher: Institute of Mathematical Statistics

    2004

  20. [28]

    A least-squares approach to direct importance estimation

    Kanamori, T., Hido, S., and Sugiyama, M. A least-squares approach to direct importance estimation. The Journal of Machine Learning Research, 10: 0 1391--1445, 2009

    2009

  21. [29]

    and Wu, J

    Karunamuni, R.J. and Wu, J. One-step minimum hellinger distance estimation. Computational statistics & data analysis, 55 0 (12): 0 3148--3164, 2011

    2011

  22. [30]

    and Ramdas, A

    Kim, I. and Ramdas, A. Dimension-agnostic inference using cross u-statistics. Bernoulli, 30 0 (1): 0 683--711, 2024

    2024

  23. [31]

    Inapparent infections and cholera dynamics

    King, A.A., Ionides, E.L., Pascual, M., and Bouma, M.J. Inapparent infections and cholera dynamics. Nature, 454 0 (7206): 0 877--880, August 2008. Number: 7206 Publisher: Nature Publishing Group

    2008

  24. [32]

    and Langer, S

    Kohler, M. and Langer, S. On the rate of convergence of fully connected deep neural network regression estimates . The Annals of Statistics, 49 0 (4): 0 2231 -- 2249, 2021

    2021

  25. [33]

    The HulC : confidence regions from convex hulls

    Kuchibhotla, A.K., Balakrishnan, S., and Wasserman, L. The HulC : confidence regions from convex hulls. Journal of the Royal Statistical Society Series B: Statistical Methodology, 86 0 (3): 0 586--622, July 2024

    2024

  26. [34]

    Cheap Bootstrap for Input Uncertainty Quantification

    Lam, H. Cheap Bootstrap for Input Uncertainty Quantification . In Proceedings of the Winter Simulation Conference , WSC '22, pages 2318--2329, Singapore, Singapore, March 2023. IEEE Press

    2023

  27. [35]

    and Rue, H

    Lenzi, A. and Rue, H. Towards black-box parameter estimation, February 2024. arXiv:2303.15041 [cs, stat]

    2024 arXiv

  28. [36]

    Neural networks for parameter estimation in intractable models

    Lenzi, A., Bessac, J., Rudi, J., and Stein, M.L. Neural networks for parameter estimation in intractable models. Computational Statistics & Data Analysis, 185: 0 107762, September 2023

    2023

  29. [37]

    Efficiency Versus Robustness : The Case for Minimum Hellinger Distance and Related Methods

    Lindsay, B.G. Efficiency Versus Robustness : The Case for Minimum Hellinger Distance and Related Methods . The Annals of Statistics, 22 0 (2): 0 1081--1114, June 1994. Publisher: Institute of Mathematical Statistics

    1994

  30. [38]

    Simulation-based Bayesian inference for epidemic models

    McKinley, T.J., Ross, J.V., Deardon, R., and Cook, A.R. Simulation-based Bayesian inference for epidemic models. Computational Statistics & Data Analysis, 71: 0 434--447, March 2014

    2014

  31. [39]

    and Retkute, R

    Minter, A. and Retkute, R. Approximate Bayesian Computation for infectious disease modelling. Epidemics, 29: 0 100368, December 2019

    2019

  32. [40]

    and Cranmer, K

    Mishra-Sharma, S. and Cranmer, K. Neural simulation-based inference approach for characterizing the Galactic Center \ ensuremath\ gamma\ \ -ray excess. Physical Review D, 105 0 (6): 0 063017, March 2022

    2022

  33. [41]

    and P \"o tscher, B.M

    Nickl, R. and P \"o tscher, B.M. Efficient simulation-based minimum distance estimation and indirect inference. Mathematical methods of statistics, 19: 0 327--364, 2010

    2010

  34. [42]

    Robust Universal Inference , July 2023

    Park, B., Balakrishnan, S., and Wasserman, L. Robust Universal Inference , July 2023

    2023

  35. [43]

    gk: An R Package for the g-and-k and generalised g-and-h Distributions , June 2017

    Prangle, D. gk: An R Package for the g-and-k and generalised g-and-h Distributions , June 2017. arXiv:1706.06889 [stat]

    2017 arXiv

  36. [44]

    and MacGillivray, H.L

    Rayner, G.D. and MacGillivray, H.L. Numerical maximum likelihood estimation for the g-and-k and generalized g-and-h distributions. Statistics and Computing, 12 0 (1): 0 57--75, January 2002

    2002

  37. [45]

    Stock and recruitment

    Ricker, W.E. Stock and recruitment. Journal of the Fisheries Board of Canada, 11 0 (5): 0 559--623, 1954

    1954

  38. [46]

    Bayesianly Justifiable and Relevant Frequency Calculations for the Applied Statistician

    Rubin, D.B. Bayesianly Justifiable and Relevant Frequency Calculations for the Applied Statistician . The Annals of Statistics, 12 0 (4): 0 1151--1172, 1984

    1984

  39. [47]

    Nonparametric regression using deep neural networks with ReLU activation function

    Schmidt-Hieber, J. Nonparametric regression using deep neural networks with ReLU activation function . The Annals of Statistics, 48 0 (4): 0 1875 -- 1897, 2020

    2020

  40. [48]

    A Permutation - Free Kernel Independence Test

    Shekhar, S., Kim, I., and Ramdas, A. A Permutation - Free Kernel Independence Test . Journal of Machine Learning Research, 24 0 (369): 0 1--68, 2023

    2023

  41. [49]

    Density ratio estimation: A comprehensive review (statistical experiment and its related topics)

    Sugiyama, M., Suzuki, T., and Kanamori, T. Density ratio estimation: A comprehensive review (statistical experiment and its related topics). Research Institute for Mathematical Sciences Kokyuroku, Kyoto University, 2010

    2010

  42. [50]

    Density Ratio Estimation in Machine Learning

    Sugiyama, M., Suzuki, T., and Kanamori, T. Density Ratio Estimation in Machine Learning . Cambridge University Press, Cambridge, 2012

    2012

  43. [51]

    and Kuchibhotla, A.K

    Takatsu, K. and Kuchibhotla, A.K. Bridging root- n and non-standard asymptotics: Adaptive inference in m-estimation. arXiv preprint arXiv:2501.07772, 2025

    2025 arXiv

  44. [52]

    Likelihood- Free Inference by Ratio Estimation

    Thomas, O., Dutta, R., Corander, J., Kaski, S., and Gutmann, M.U. Likelihood- Free Inference by Ratio Estimation . Bayesian Analysis, 17 0 (1): 0 1--31, March 2022. Publisher: International Society for Bayesian Analysis

    2022

  45. [53]

    Asymptotic Statistics

    Vaart, A.W.V.D. Asymptotic Statistics . Cambridge University Press, 1 edition, October 1998

    1998

  46. [54]

    and Wasserman, L

    Verdinelli, I. and Wasserman, L. Decorrelated Variable Importance . Journal of Machine Learning Research, 25 0 (7): 0 1--27, 2024

    2024

  47. [55]

    Neural likelihood surfaces for spatial processes with computationally intensive or intractable likelihoods

    Walchessen, J., Lenzi, A., and Kuusela, M. Neural likelihood surfaces for spatial processes with computationally intensive or intractable likelihoods. Spatial Statistics, 62: 0 100848, August 2024

    2024

  48. [56]

    and Nowak, R.D

    Willett, R.M. and Nowak, R.D. Minimax Optimal Level - Set Estimation . IEEE Transactions on Image Processing, 16 0 (12): 0 2965--2979, December 2007. Conference Name: IEEE Transactions on Image Processing

    2007

  49. [57]

    and Wang, P

    Xie, M.g. and Wang, P. Repro Samples Method for Finite - and Large - Sample Inferences , June 2022. arXiv:2206.06421

    2022 arXiv

  50. [58]

    and Yao, W

    Zhao, Z. and Yao, W. Sequential design for nonparametric inference. Canadian Journal of Statistics, 40 0 (2): 0 362--377, 2012

    2012

Pith tools

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