pith. sign in

arxiv: 2302.14119 · v10 · pith:WIA3DE5Anew · submitted 2023-02-27 · 🧮 math.NA · cs.NA

Double-loop randomized quasi-Monte Carlo estimator for nested integration

Arved Bartuska , Andr\'e Gustavo Carlon , Luis Espath , Sebastian Krumscheid , Ra\'ul Tempone This is my paper

Pith reviewed 2026-05-24 09:31 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords nested integrationrandomized quasi-Monte Carloerror boundsbias and varianceOwen scramblingexpected information gainnumerical integration
0
0 comments X

The pith

A double-loop randomized quasi-Monte Carlo estimator yields asymptotic bounds on bias and variance for nested integrals.

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

The paper introduces a nested randomized quasi-Monte Carlo method that approximates both the inner integral and the outer integral of the form integral f of inner integral at the same time. It derives asymptotic bounds on the bias and variance of the resulting estimator, together with the regularity conditions needed for those bounds to hold. The approach incorporates Owen scrambling to cover integrands that have infinite Hardy-Krause variation and supplies a truncation device for expected-information-gain calculations. The goal is to obtain a cheaper alternative to ordinary nested Monte Carlo sampling for problems in engineering and mathematical finance.

Core claim

The central claim is that the proposed double-loop rQMC estimator admits asymptotic error bounds for both bias and variance, attainable once the integrands meet the stated regularity conditions, including the conditions that permit Owen scrambling for functions of infinite Hardy-Krause variation.

What carries the argument

The double-loop randomized quasi-Monte Carlo estimator that couples rQMC sampling across the inner and outer integrals of a nested problem.

If this is right

  • Bias and variance of the estimator obey explicit asymptotic rates once the regularity conditions are met.
  • Owen scrambling extends the error analysis to integrands whose Hardy-Krause variation is infinite.
  • A truncation scheme produces a practical estimator for expected information gain.
  • The method improves computational efficiency relative to standard nested Monte Carlo sampling.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the derived rates are observed in practice, the number of inner-loop samples needed for a target accuracy could drop substantially compared with plain Monte Carlo.
  • The same double-loop structure might be reused for other nonlinear nested expectations that appear in stochastic optimization or Bayesian experimental design.

Load-bearing premise

The integrands must satisfy the regularity conditions, including those that allow Owen scrambling, that are required for the asymptotic error bounds to hold.

What would settle it

A concrete numerical experiment in which the observed decay rates of bias or variance fail to match the predicted asymptotic rates when the integrands are known to obey the stated regularity conditions.

Figures

Figures reproduced from arXiv: 2302.14119 by Andr\'e Gustavo Carlon, Arved Bartuska, Luis Espath, Ra\'ul Tempone, Sebastian Krumscheid.

Figure 1
Figure 1. Figure 1: Example 1: Optimal number of the outer (N∗ ) and inner (M∗ ) samples vs. toler￾ance (T OL) for the rDLQMCIS, DLMCIS, DLMC, rQMCLA, and MCLA estimators. Only one inner sample is required for high tolerances in the rDLQMCIS and DLMCIS estimators due to importance sampling. Hence, the projected rate is attained only for low tolerances. No inner sampling is necessary for the rQMCLA and MCLA estimators; however… view at source ↗
Figure 2
Figure 2. Figure 2: demonstrates the optimal work for the rDLQMCIS, DLMCIS, DLMC, rQMCLA, and MCLA estimators. Although the latter two are highly efficient, the inherent bias from the Laplace approximation renders them ineffective for low tolerances. 10−6 10−5 10−4 10−3 10−2 10−1 100 T OL 103 107 1011 1015 1019 1023 W ork = N∗ × M∗ rDLQMCIS rate 1.73 DLMCIS rate 3 DLMC rate 3 rQMCLA rate 1.08 MCLA rate 2 [PITH_FULL_IMAGE:fig… view at source ↗
Figure 3
Figure 3. Figure 3: Example 2: Rectangular domain with the circular exclusion and sensor locations {(0.0 + ξ1, 1.0), (0.2 + ξ1, 1.0) (0.4 + ξ1, 1.0) (0.6 + ξ1, 1.0)}, where green indicates strain and red indicates temperature sensors) [PITH_FULL_IMAGE:figures/full_fig_p025_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Example 2: Temperature increases at observation times of the experiment for the rectangular domain with a circular exclusion. Point-sensor locations are displayed as circles at the top of the domain. 10−6 10−5 10−4 10−3 10−2 10−1 100 T OL 102 104 106 108 1010 1012 1014 M∗ , N∗ rDLQMC: N∗ rate 1.25 DLMC: N∗ rate 2 rDLQMC: M∗ rate 0.69 DLMC: M∗ rate 1 [PITH_FULL_IMAGE:figures/full_fig_p026_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Example 2: Optimal number of the outer (N∗ ) and inner (M∗ ) samples vs. tol￾erance (T OL) for the rDLQMC and DLMC estimators [PITH_FULL_IMAGE:figures/full_fig_p026_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Example 2: Optimal splitting parameter (κ ∗ ) vs. tolerance T OL for the rDLQMC, and DLMC estimators. 10−6 10−5 10−4 10−3 10−2 10−1 100 T OL 106 1010 1014 1018 1022 1026 1030 W ork = N∗ × M∗ × h∗−γ rDLQMC rate 1.93+1.08 DLMC rate 3+1.08 [PITH_FULL_IMAGE:figures/full_fig_p027_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Example 2: Optimal work (N∗×M∗×h ∗−γ ) vs. tolerance T OL for the rDLQMC and DLMC estimators [PITH_FULL_IMAGE:figures/full_fig_p027_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Example 2: Expected information gain (EIG) as a function of the design ξ = (ξ1, ξ2), indicating sensor locations and observation times, estimated using the rDLQMC estimator with an allowed tolerance of T OL = 0.2. The maximum EIG is reached for ξ1 = 0.4 and ξ2 = 8 × 105 . 20000 ξ2 40000 60000 80000 ξ1 0.0 0.1 0.2 0.3 0.4 EIG 2.3 2.4 2.5 2.6 2.7 2.4 2.5 2.6 2.7 [PITH_FULL_IMAGE:figures/full_fig_p028_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Example 2: Expected information gain (EIG) as a function of the design ξ = (ξ1, ξ2), indicating sensor locations and observation times, estimated using the rDLQMC estimator with an allowed tolerance of T OL = 0.2. The maximum EIG is reached for ξ1 = 0.4 and ξ2 = 8 × 105 [PITH_FULL_IMAGE:figures/full_fig_p028_9.png] view at source ↗
read the original abstract

Nested integration of the form $\int f\left(\int g(\bs{y},\bs{x})\di{}\bs{x}\right)\di{}\bs{y}$, characterized by an outer integral connected to an inner integral through a nonlinear function $f$, is a challenging problem in various fields, such as engineering and mathematical finance. The available numerical methods for nested integration based on Monte Carlo (MC) methods can be prohibitively expensive owing to the error propagating from the inner to the outer integral. Attempts to enhance the efficiency of these approximations using the quasi-MC (QMC) or randomized QMC (rQMC) method have focused on either the inner or outer integral approximation. This work introduces a novel nested rQMC method that simultaneously addresses the approximation of the inner and outer integrals. The method leverages the unique nested integral structure to offer a more efficient approximation mechanism. As the primary contribution, we derive asymptotic error bounds for the bias and variance of our estimator, along with the regularity conditions under which these bounds can be attained. Incorporating Owen's scrambling techniques, we address integrands exhibiting infinite variation in the Hardy--Krause sense, enabling theoretically sound error estimates. Moreover, we derive a truncation scheme for applications in the context of expected information gain estimation. We verify the estimator quality through numerical experiments by comparing the computational efficiency of the nested rQMC method against standard nested MC estimation to highlight the computational savings and enhanced applicability of the proposed approach.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 3 minor

Summary. The manuscript introduces a double-loop randomized quasi-Monte Carlo (rQMC) estimator for nested integrals of the form ∫ f(∫ g(y,x) dx) dy. The central contribution is the derivation of asymptotic bounds on the bias and variance of the estimator together with the precise regularity conditions (including Owen scrambling to accommodate infinite Hardy-Krause variation) under which the bounds are attained. A truncation scheme for expected information gain estimation is also derived, and numerical experiments compare the computational efficiency of the new estimator against standard nested Monte Carlo.

Significance. If the derived asymptotic bounds hold under the stated regularity conditions, the work supplies a theoretically justified and more efficient alternative to nested Monte Carlo for a class of integrals that arise in engineering and mathematical finance. The explicit treatment of infinite-variation integrands via Owen scrambling and the truncation scheme for information-gain applications are concrete strengths that enhance both the theoretical scope and practical utility of the estimator.

minor comments (3)
  1. [§3] §3 (or the theorem stating the main bounds): the precise rates (e.g., O(N^{-1+ε}) or similar) for bias and variance should be written explicitly rather than left as “asymptotic error bounds.”
  2. [Numerical experiments] The numerical experiments section would benefit from a table or plot that isolates the contribution of the inner-loop rQMC versus the outer-loop rQMC to the observed error reduction.
  3. A short remark clarifying whether the truncation scheme for expected information gain preserves the same asymptotic rates as the untruncated estimator would remove ambiguity for readers applying the method in that setting.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the constructive and positive report, which correctly identifies the main contributions of the paper: the derivation of asymptotic bias and variance bounds for the double-loop rQMC estimator under the stated regularity conditions (including Owen scrambling for infinite Hardy-Krause variation) and the truncation scheme for expected information gain. We are pleased that the referee views the work as supplying a theoretically justified and more efficient alternative to nested Monte Carlo. Since the report contains no specific major comments, we address the recommendation of minor revision below and will incorporate any editorial or minor clarifications in the revised manuscript.

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper derives asymptotic bias and variance bounds for its double-loop rQMC estimator directly from the estimator definition together with stated regularity conditions on the integrands (including the extension of Owen scrambling to infinite Hardy-Krause variation). These bounds are obtained by standard QMC error analysis and are not obtained by fitting parameters to data subsets, by renaming an input quantity as a prediction, or by any self-referential definition. The truncation scheme for expected information gain is presented as a separate, independent contribution. No load-bearing step reduces to a self-citation chain, an imported uniqueness theorem, or an ansatz smuggled via prior work by the same authors. The derivation chain is therefore self-contained against external mathematical benchmarks and the paper's own assumptions.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claims rest on domain assumptions about integrand regularity that are standard in QMC theory but not independently verified in the provided abstract.

axioms (1)
  • domain assumption Integrands satisfy the regularity conditions required for the asymptotic error bounds and for Owen's scrambling to apply to infinite-variation cases.
    Explicitly invoked in the abstract as necessary for the derived bounds to hold.

pith-pipeline@v0.9.0 · 5805 in / 1098 out tokens · 32688 ms · 2026-05-24T09:31:09.755068+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

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

  1. Optimal Experimental Design for Reliable Learning of History-Dependent Constitutive Laws

    cond-mat.mtrl-sci 2026-03 unverdicted novelty 6.0

    A Bayesian optimal experimental design framework with Gaussian approximation of expected information gain and surrogate Fisher information enables optimized uniaxial tests that significantly improve identifiability of...

Reference graph

Works this paper leans on

65 extracted references · 65 canonical work pages · cited by 1 Pith paper · 1 internal anchor

  1. [1]

    Estimating expected information gains for experimental designs with application to the random fatigue-limit model

    Ryan KJ. Estimating expected information gains for experimental designs with application to the random fatigue-limit model. Journal of Computational and Graphical Statistics , 12:585–603, 2003

    work page 2003

  2. [2]

    Fast Bayesian experimental design: Laplace-based importance sampling for the expected information gain

    Beck J, Dia BM, Espath LF, Long Q, and Tempone R. Fast Bayesian experimental design: Laplace-based importance sampling for the expected information gain. Computer Methods in Applied Mechanics and Engineering , 334:523–553, 2018

    work page 2018

  3. [3]

    Multilevel double loop Monte Carlo and stochastic collocation methods with importance sampling for Bayesian optimal experimental design

    Beck J, Dia BM, Espath LF, and Tempone R. Multilevel double loop Monte Carlo and stochastic collocation methods with importance sampling for Bayesian optimal experimental design. International Journal for Numerical Methods in Engineering, 121:3482–3503, 2020

    work page 2020

  4. [4]

    Nesterov-aided stochastic gradient methods using Laplace approximation for Bayesian design optimization

    Carlon AG, Dia BM, Espath LF, Lopez RH, and Tempone R. Nesterov-aided stochastic gradient methods using Laplace approximation for Bayesian design optimization. Computer Methods in Applied Mechanics and Engineering , 363:112909, 2020

    work page 2020

  5. [5]

    A mathematical theory of communication

    Shannon CE. A mathematical theory of communication. Bell System Technical Journal , 27:379–423, 1948

    work page 1948

  6. [6]

    Information theory and statistics

    Kullback S. Information theory and statistics . Wiley, 1959. RANDOMIZED QUASI-MONTE CARLO ESTIMATOR FOR NESTED INTEGRATION 39

    work page 1959

  7. [7]

    Fast estimation of expected information gains for Bayesian experimental designs based on Laplace approximations

    Long Q, Scavino M, Tempone R, and Wang S. Fast estimation of expected information gains for Bayesian experimental designs based on Laplace approximations. Computer Methods in Applied Mechanics and Engineering , 259:24–39, 2013

    work page 2013

  8. [8]

    Multimodal information gain in Bayesian design of experiments

    Long Q. Multimodal information gain in Bayesian design of experiments. arXiv preprint arXiv:2108.07224 , 2021

    work page arXiv 2021

  9. [9]

    On the convergence of the Laplace approximation and noise-level-robustness of Laplace-based Monte Carlo methods for Bayesian inverse problems

    Schillings C, Sprungk B, and Wacker P. On the convergence of the Laplace approximation and noise-level-robustness of Laplace-based Monte Carlo methods for Bayesian inverse problems. Numerische Mathematik, 145:915–971, 2020

    work page 2020

  10. [10]

    Laplace's method in Bayesian inverse problems

    Wacker P. Laplace’s method in Bayesian inverse problems. arXiv preprint arXiv:1701.07989 , 2017

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  11. [11]

    Bayesian experimental design: a review

    Chaloner K and Verdinelli I. Bayesian experimental design: a review. Statistical Sciences, 10:273–304, 1995

    work page 1995

  12. [12]

    On information and sufficiency

    Kullback S and Leibler RA. On information and sufficiency. Annals of Mathematical Statistics , 22:79–86, 1951

    work page 1951

  13. [13]

    Laplace’s 1774 memoir on inverse probability

    Stigler SM. Laplace’s 1774 memoir on inverse probability. Statistical Science, 1:359–378, 1986

    work page 1986

  14. [14]

    On a measure of information provided by an experiment

    Lindley DV. On a measure of information provided by an experiment. Annals of Mathematical Statistics , 27:986–1005, 1956

    work page 1956

  15. [15]

    Accurate approximations for posterior moments and marginal densities.Journal of the American Statistical Association, 81:82–86, 1986

    Tierney L and Kadane JB. Accurate approximations for posterior moments and marginal densities.Journal of the American Statistical Association, 81:82–86, 1986

    work page 1986

  16. [16]

    Fully exponential Laplace approximations to expectations and variances of nonpositive functions

    Tierney L, Kass RE, and Kadane JB. Fully exponential Laplace approximations to expectations and variances of nonpositive functions. Journal of the American Statistical Association , 84:710–716, 1989

    work page 1989

  17. [17]

    The validity of posterior expansions based on Laplace’s method

    Kass RE, Tierney L, and Kadane JB. The validity of posterior expansions based on Laplace’s method. in: Geisser S, Hodges JS, Press SJ, and Zellner A (Eds.), Essays in Honor of George Barnard , 473–488, 1990

    work page 1990

  18. [18]

    A Laplace method for under-determined Bayesian optimal experimental designs

    Long Q, Scavino M, Tempone R, and Wang S. A Laplace method for under-determined Bayesian optimal experimental designs. Computer Methods in Applied Mechanics and Engineering , 285:849–876, 2015

    work page 2015

  19. [19]

    Multilevel Monte Carlo path simulation

    Giles MB. Multilevel Monte Carlo path simulation. Operations Research, 56(3):607–617, 2008

    work page 2008

  20. [20]

    Efficient Bayesian experimentation using an expected information gain lower bound

    Tsilifis P, Ghanem RG, and Hajali P. Efficient Bayesian experimentation using an expected information gain lower bound. SIAM/ASA Journal on Uncertainty Quantification , 5:30–62, 2017

    work page 2017

  21. [21]

    Decision-theoretic sensitivity analysis for reservoir development under uncertainty using multilevel quasi-Monte Carlo methods

    Goda T, Daisuke M, Kei T, and Kozo S. Decision-theoretic sensitivity analysis for reservoir development under uncertainty using multilevel quasi-Monte Carlo methods. Computational Geosciences, 22:1009–1020, 2018

    work page 2018

  22. [22]

    Efficient risk estimation via nested multilevel quasi-Monte Carlo simulation

    Xu Z, He Z, and Wang X. Efficient risk estimation via nested multilevel quasi-Monte Carlo simulation. Journal of Com- putational and Applied Mathematics , 443:115745, 2024

    work page 2024

  23. [23]

    Multilevel and quasi Monte Carlo methods for the calculation of the expected value of partial perfect information

    Fang W, Wang Z, Giles MB, Jackson CH, Welton NJ, Andrieu C, and Thom H. Multilevel and quasi Monte Carlo methods for the calculation of the expected value of partial perfect information. Medical Decision Making, 42(2):168–181, 2022

    work page 2022

  24. [24]

    Monte Carlo and quasi-Monte Carlo methods

    Caflisch RE. Monte Carlo and quasi-Monte Carlo methods. Acta Numerica, 1–49, 1998

    work page 1998

  25. [25]

    Lattice rules: How well do they measure up? In: Hellekalek P and Larcher G (Eds.)

    Hickernell FJ. Lattice rules: How well do they measure up? In: Hellekalek P and Larcher G (Eds.). Random and Quasi- Random Point Sets. Lecture Notes in Statistics 138 , New York, Springer, 109–166, 1998

    work page 1998

  26. [26]

    Random number generation and quasi-Monte Carlo methods

    Niederreiter H. Random number generation and quasi-Monte Carlo methods. SIAM, Philadelphia, 1992

    work page 1992

  27. [27]

    Quasi-Monte Carlo sampling

    Owen AB. Quasi-Monte Carlo sampling. In: Monte Carlo Ray Tracing: SIGGRAPH , 69–88, 2003

    work page 2003

  28. [28]

    Randomization of quasi-Monte Carlo methods for error estimation: survey and normal approximation

    Tuffin B. Randomization of quasi-Monte Carlo methods for error estimation: survey and normal approximation. In: Monte Carlo Methods and Applications , De Gruyter, 10(3-4):617–628, 2004

    work page 2004

  29. [29]

    Probability: Theory and examples

    Durrett R. Probability: Theory and examples. Cambridge University Press , 2019

    work page 2019

  30. [30]

    Digital nets and sequences: discrepancy theory and quasi-Monte Carlo integration.Cambridge University Press, 2010

    Dick J and Pillichshammer F. Digital nets and sequences: discrepancy theory and quasi-Monte Carlo integration.Cambridge University Press, 2010

    work page 2010

  31. [31]

    High-dimensional integration: the quasi-Monte Carlo way

    Dick J, Kuo FY, and Sloan IH. High-dimensional integration: the quasi-Monte Carlo way. Acta Numerica, 133–288, 2013

    work page 2013

  32. [32]

    Distribution of points in a cube and approximate evaluation of integrals

    Sobol’ IM. Distribution of points in a cube and approximate evaluation of integrals. ˇZ. Vyˇ cisl. Mat. i Mat. Fiz. (in Russian), 7:784–802, 1967

    work page 1967

  33. [33]

    Improving the efficiency of fully Bayesian optimal design of experiments using randomised Quasi-Monte Carlo

    Drovandi CC and Tran M-N. Improving the efficiency of fully Bayesian optimal design of experiments using randomised Quasi-Monte Carlo. Bayesian Analysis, 13(1):139–162, 2018

    work page 2018

  34. [34]

    On the distribution of integration error by randomly-shifted lattice rules

    L’Ecuyer P, Munger D, and Tuffin B. On the distribution of integration error by randomly-shifted lattice rules. Electronic Journal of Statistics , 4:950–993, 2010

    work page 2010

  35. [35]

    Mean estimation for randomized Quasi Monte Carlo method

    Gobet E, Lerasle M, and M´ etivier D. Mean estimation for randomized Quasi Monte Carlo method. Hal preprint hal- 03631879v2, 2022

    work page 2022

  36. [36]

    Non-asymptotic error estimates for the Laplace approximation in Bayesian inverse problems

    Helin T and Kretschmann R. Non-asymptotic error estimates for the Laplace approximation in Bayesian inverse problems. Numerische Mathematik, 150:521–549, 2022

    work page 2022

  37. [37]

    Randomized Quasi-Monte Carlo: An introduction for practitioners

    L’Ecuyer P. Randomized Quasi-Monte Carlo: An introduction for practitioners. In: Owen A and Glynn P (Eds.). Monte Carlo and Quasi-Monte Carlo Methods. MCQMC 2016. Springer Proceedings in Mathematics & Statistics , Cham, Springer 241:29–52, 2018

    work page 2016

  38. [38]

    Funktionen von beschr¨ ankter Variation in der Theorie der Gleichverteilung

    Hlawka E. Funktionen von beschr¨ ankter Variation in der Theorie der Gleichverteilung. Annali di Matematica Pura ed Applicata, 54:325–333, 1961

    work page 1961

  39. [39]

    Local antithetic sampling with scrambled nets

    Owen AB. Local antithetic sampling with scrambled nets. The Annals of Statistics , 36(5):2319–2343, 2008

    work page 2008

  40. [40]

    An unconditionally stable staggered algorithm for transient finite element analysis of coupled thermoelastic problems

    Farhat C, Park KC, and Dubois-Pelerin Y. An unconditionally stable staggered algorithm for transient finite element analysis of coupled thermoelastic problems. Computer Methods in Applied Mechanics and Engineering , 85(3):349–365, 1991

    work page 1991

  41. [41]

    Multi-index stochastic collocation convergence rates for random PDEs with parametric regularity

    Haji-Ali A-L, Nobile F, Tamellini, L, and Tempone R. Multi-index stochastic collocation convergence rates for random PDEs with parametric regularity. Foundations of Computational Mathematics , 16:1555–1605, 2016

    work page 2016

  42. [42]

    Halton sequences avoid the origin

    Owen AB. Halton sequences avoid the origin. SIAM Review, 48:487–503, 2006

    work page 2006

  43. [43]

    Nonasymptotic convergence rate of quasi-Monte Carlo: applications to linear elliptic PDEs with lognormal coefficients and importance samplings

    Liu Y and Tempone R. Nonasymptotic convergence rate of quasi-Monte Carlo: applications to linear elliptic PDEs with lognormal coefficients and importance samplings. arXiv preprint arXiv:2310.14351 , 2023

    work page arXiv 2023

  44. [44]

    Quasi-Monte Carlo finite element methods for elliptic PDEs with lognormal random coefficients

    Graham IG, Kuo FY, Nichols JA, Scheichl R, Schwab C, and Sloan IH. Quasi-Monte Carlo finite element methods for elliptic PDEs with lognormal random coefficients. Numerische Mathematik, 131:329–368, 2015. 40 RANDOMIZED QUASI-MONTE CARLO ESTIMATOR FOR NESTED INTEGRATION

    work page 2015

  45. [45]

    On the error rate of importance sampling with randomized quasi-Monte Carlo

    He Z, Zheng Z, and Wang X. On the error rate of importance sampling with randomized quasi-Monte Carlo. SIAM Journal on Numerical Analysis, 61(2):10.1137/22M1510121, 2023

    work page doi:10.1137/22m1510121 2023

  46. [46]

    Quasi-Monte Carlo for unbounded integrands with importance sampling

    Ouyang D, Wang X, and He Z. Quasi-Monte Carlo for unbounded integrands with importance sampling. arXiv preprint arXiv:2310.00650, 2023

    work page arXiv 2023

  47. [47]

    Randomly permuted (t,m,s)-nets and (t,s)-sequences

    Owen AB. Randomly permuted (t,m,s)-nets and (t,s)-sequences. in: Niederreiter H and Shiue PJS (Eds.), Monte Carlo and Quasi-Monte Carlo methods in Scientific Computing. Lecture Notes in Statistics 106 , New York, Springer, 1995

    work page 1995

  48. [48]

    On nesting Monte Carlo estimators

    Rainforth T, Cornish R, Yang H, Warrington A, and Wood F. On nesting Monte Carlo estimators. Proceedings of the 35th International Conference on Machine Learning, PLMR 80 Stockholm, 2008

    work page 2008

  49. [49]

    Estimating the mean of a non-linear function of conditional expectation

    Hong LJ, and Juneja S. Estimating the mean of a non-linear function of conditional expectation. in: Rossetti MD, Hill RR, Johansson B, Dunkin A and Ingalls RG (Eds.), Proceedings of the 2009 Winter Simulation Conference (WSC) Austin, 1223–1236, 2009

    work page 2009

  50. [50]

    Confidence intervals for randomized quasi-Monte Carlo estimators

    L’Ecuyer P, Nakayama MK, Owen AB, and Tuffin B. Confidence intervals for randomized quasi-Monte Carlo estimators. Winter Simulation Conference, San Antonio, US , 2023

    work page 2023

  51. [51]

    MCMC design-based non-parametric regression for rare event

    Fort G, Gobet E, and Moulines E. MCMC design-based non-parametric regression for rare event. Application to nested risk computations. Monte Carlo Methods and Applications , 23(1):21–42, 2017

    work page 2017

  52. [52]

    Variational Bayesian optimal experimental design

    Foster A, Jankowiak M, Bingham E, Horsfall P, Teh YW, Rainforth T, and Goodman N. Variational Bayesian optimal experimental design. Advances in Neural Information Processing Systems 32 , 2019

    work page 2019

  53. [53]

    Multilevel Monte Carlo estimation of expected information gains.Stochastic Analysis and Applications, 38(4):581–600, 2020

    Goda T, Hironaka T, and Iwamoto T. Multilevel Monte Carlo estimation of expected information gains.Stochastic Analysis and Applications, 38(4):581–600, 2020

    work page 2020

  54. [54]

    Unbiased MLMC stochastic gradient-based optimization of Bayesian experimental designs

    Goda T, Hironaka T, Kitade W, and Foster A. Unbiased MLMC stochastic gradient-based optimization of Bayesian experimental designs. SIAM Journal on Scientific Computing , 44(1):A286–A311, 2022

    work page 2022

  55. [55]

    Quasi-Monte Carlo finite element methods for a class of elliptic partial differential equations with random coefficients

    Kuo FY, Schwab C, and Sloan IH. Quasi-Monte Carlo finite element methods for a class of elliptic partial differential equations with random coefficients. SIAM Journal on Numerical Analysis , 50(6):3351–3374, 2012

    work page 2012

  56. [56]

    Multi-level quasi-Monte Carlo finite element methods for a class of elliptic PDEs with random coefficients

    Kuo FY, Schwab C, and Sloan IH. Multi-level quasi-Monte Carlo finite element methods for a class of elliptic PDEs with random coefficients. Foundations of Computational Mathematics , 15:411–449, 2015

    work page 2015

  57. [57]

    Towards Bayesian experimental design for nonlinear models that require a large number of sampling times

    Ryan EG, Drovandi CC, Thompson MH and Pettitt AN. Towards Bayesian experimental design for nonlinear models that require a large number of sampling times. Computational Statistics and Data Analysis , 70:45–60, 2014

    work page 2014

  58. [58]

    Scrambled net variance for integrals of smooth functions

    Owen AB. Scrambled net variance for integrals of smooth functions. The Annals of Statistics , 25:1541–1562, 1997

    work page 1997

  59. [59]

    Randomization of number theoretic methods for multiple integration

    Cranley R, and Patterson TNL. Randomization of number theoretic methods for multiple integration. SIAM Journal on Numerical Analysis, 13(6):904–914, 1976

    work page 1976

  60. [60]

    The nonzero gain coefficients of Sobol’s sequences are always powers of two

    Pan Z, and Owen AB. The nonzero gain coefficients of Sobol’s sequences are always powers of two. Journal of Complexity , 75:101700, 2023

    work page 2023

  61. [61]

    Approximating probability densities by iterated Laplace approximations

    Bornkamp B. Approximating probability densities by iterated Laplace approximations. Journal of Computational and Graphical Statistics, 20(3):656–669, 2011

    work page 2011

  62. [62]

    Optimal Bayesian experimental design for priors of compact sup- port with application to shock-tube experiments for combustion kinetics

    Bisetti F, Kim D, Knio O, Long Q, and Tempone R. Optimal Bayesian experimental design for priors of compact sup- port with application to shock-tube experiments for combustion kinetics. International Journal for Numerical Methods in Engineering, 108:136–155, 2016

    work page 2016

  63. [63]

    Quasi-Monte Carlo for Bayesian design of experiment problems governed by parametric PDEs

    Kaarnioja V, and Schillings C. Quasi-Monte Carlo for Bayesian design of experiment problems governed by parametric PDEs. arXiv preprint arXiv:2405.03529 [v2] , 2024

    work page arXiv 2024

  64. [64]

    Multimodal information gain in Bayesian design of experiments

    Long Q. Multimodal information gain in Bayesian design of experiments. Computational Statistics, 37(2):865–885 2022

    work page 2022

  65. [65]

    Handbook of the normal distribution

    Patel JK, and Read CB. Handbook of the normal distribution. 2nd ed., New York, Marcel Dekker, 1996

    work page 1996