pith. machine review for the scientific record. sign in

arxiv: 2604.06417 · v1 · submitted 2026-04-07 · 📊 stat.CO

Recognition: no theorem link

Niching Importance Sampling for Multi-modal Rare-event Simulation

Hugh J. Kinnear , F.A. DiazDelaO
Authors on Pith no claims yet

Pith reviewed 2026-05-10 17:52 UTC · model grok-4.3

classification 📊 stat.CO
keywords nichingimportance samplingrare-event simulationreliability analysismulti-modalprobability of failurecross-entropy minimizationMarkov chain
0
0 comments X

The pith

Niching importance sampling combines evolutionary diversity maintenance with importance sampling and cross-entropy updates to produce stable estimates of failure probability on multi-modal performance functions.

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

The paper proposes niching importance sampling as a way to estimate small probabilities of failure when the underlying performance function has multiple modes. Standard reliability methods using importance sampling often collapse or produce high variance in these geometries because samples concentrate in only one mode. The new framework draws niching ideas from evolutionary optimization to keep sampling diversity across modes while still using Markov chains and relative cross-entropy minimization to refine the importance distribution. A sympathetic reader would care because rare-event probabilities appear in safety-critical engineering calculations, and existing tools become unreliable precisely when the failure landscape is most complicated. If the approach works as described, it yields estimators that remain effective without sacrificing the variance reduction that makes importance sampling useful in the first place.

Core claim

Niching importance sampling is a framework that merges Markov chains, importance sampling, and relative cross-entropy minimization from reliability analysis with niching techniques from evolutionary multi-modal optimization. The result is a robust estimator of the probability of failure that avoids the degenerate behavior observed in existing methods when applied to several multi-modal performance functions.

What carries the argument

Niching importance sampling, which maintains population diversity across modes during the sampling and distribution-update steps while preserving the unbiasedness and variance-reduction properties of importance sampling.

If this is right

  • The method produces consistent estimates across a range of numerical examples that cause degeneracy in standard reliability techniques.
  • The estimator remains unbiased while retaining the efficiency gains of importance sampling.
  • It handles performance functions whose failure regions consist of disconnected components without requiring prior knowledge of mode locations.
  • The approach scales to problems where the geometry of the limit-state surface is complex rather than simple.

Where Pith is reading between the lines

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

  • The same niching logic could be paired with other variance-reduction techniques such as subset simulation when modes are separated by high-probability barriers.
  • Because the update step still relies on cross-entropy, the method may inherit sensitivity to the choice of parametric family for the importance density.
  • In applications where the true failure probability is unknown, the method could be paired with a cheap pilot run to detect the number of modes before full sampling begins.

Load-bearing premise

That niching can be added to importance sampling and relative cross-entropy minimization without introducing bias or losing the variance-reduction benefits on the underlying reliability geometry.

What would settle it

On one of the multi-modal test functions used in the paper, the estimator either produces a variance that grows with sample size or returns a value far from the known true probability of failure.

Figures

Figures reproduced from arXiv: 2604.06417 by F.A. DiazDelaO, Hugh J. Kinnear.

Figure 1
Figure 1. Figure 1: Solution trajectories of points sampled from the input distribution. [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Degenerate SIS runs on performance functions with challenging topology. [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Kernel density estimate of the SIS estimator for the piecewise linear function. [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Chain runs of niching initial sampling procedure for challenging performance functions. [PITH_FULL_IMAGE:figures/full_fig_p014_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Markov chains of the niching importance sampling procedure after one update, applied to challenging performance [PITH_FULL_IMAGE:figures/full_fig_p015_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Importance samples generated by the importance distribution fit using niching importance sampling for two chal [PITH_FULL_IMAGE:figures/full_fig_p019_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Two-degree-of-freedom mass spring system. [PITH_FULL_IMAGE:figures/full_fig_p024_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Passive vehicle suspension model. Dimension Method Mean PF CoV PF Mean g evals 3 SIS 0 - 1.65 × 104 iCE 1.32 × 10−6 0.25 1.14 × 104 NIS 1.42 × 10−6 0.04 2.22 × 103 99 SIS 0 - 1.67 × 104 iCE 1.25 × 10−8 3.82 9.28 × 103 NIS 1.28 × 10−6 0.10 1.17 × 104 300 SIS 1.52 × 10−8 6.52 1.64 × 104 iCE 5.36 × 10−11 9.91 1.21 × 104 NIS 1.33 × 10−6 0.12 2.84 × 104 [PITH_FULL_IMAGE:figures/full_fig_p025_8.png] view at source ↗
read the original abstract

This paper proposes niching importance sampling, a framework that combines concepts from reliability analysis, e.g. Markov chains, importance sampling, and relative cross entropy minimisation, with niching techniques from evolutionary multi-modal optimisation. The result is a highly robust estimator of the probability of failure, that can tackle sampling challenges posed by the underlying geometry of a reliability problem. Niching importance sampling is tested on a range of numerical examples and is shown to consistently avoid the degenerate behaviour observed for existing reliability methods on several multi-modal performance functions.

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

3 major / 2 minor

Summary. The paper proposes niching importance sampling, a hybrid method that augments standard importance sampling and relative cross-entropy minimization with niching operators (clustering and per-niche resampling) drawn from evolutionary multi-modal optimization. The central claim is that the resulting estimator remains robust for multi-modal failure domains, consistently avoiding the degeneracy that standard reliability methods exhibit on the tested performance functions.

Significance. If the niching step can be shown to preserve unbiasedness and the variance-reduction properties of importance sampling, the framework would address a recognized practical limitation in rare-event simulation for reliability problems whose limit-state surfaces contain multiple disconnected modes. The numerical examples provide initial evidence of improved stability, but the absence of a formal proof that the mixture proposal remains optimal for the global cross-entropy objective limits the immediate theoretical impact.

major comments (3)
  1. [§3.2] §3.2 (Niching operator and per-niche CE minimization): the manuscript must demonstrate that the overall estimator E[I(g(X)≤0) w(X)] remains exactly unbiased after the clustering-plus-resampling step. If niche weights or proposal parameters are chosen by an evolutionary heuristic rather than by a single global minimization of the KL divergence to the optimal importance density, the resulting mixture is no longer the solution of the original cross-entropy problem and a non-zero bias can appear; a short proof or explicit statement that the expectation is unchanged is required.
  2. [§4] §4 (Numerical examples): the reported failure-probability estimates for the multi-modal test cases lack accompanying standard-error or effective-sample-size diagnostics that would allow the reader to verify that the observed stability is not purchased at the cost of increased variance. Adding these quantities (or at least the coefficient of variation) for both the proposed method and the baseline IS/CE estimators would strengthen the empirical claim.
  3. [§2.3] §2.3 (Relative cross-entropy minimization): the paper should clarify whether the per-niche CE problems are solved independently or whether a global normalization of the mixture weights is performed after niching. The current description leaves open the possibility that the final proposal is no longer the minimizer of the single global KL objective, which would invalidate the usual optimality guarantees.
minor comments (2)
  1. Notation for the niching operator (e.g., the clustering function C and the resampling weights) is introduced without a compact mathematical definition; a single displayed equation summarizing the modified sampling measure would improve readability.
  2. The abstract states that the method 'consistently avoid[s] the degenerate behaviour,' but the main text should explicitly define what constitutes degeneracy (e.g., effective sample size below a threshold, or variance explosion) so that the comparison with existing methods is unambiguous.

Simulated Author's Rebuttal

3 responses · 1 unresolved

We thank the referee for the careful reading and constructive major comments. We address each point below and have revised the manuscript to strengthen the presentation of unbiasedness, add variance diagnostics, and clarify the algorithmic details. The revisions preserve the original claims while improving rigor and readability.

read point-by-point responses

  1. Referee: [§3.2] the manuscript must demonstrate that the overall estimator E[I(g(X)≤0) w(X)] remains exactly unbiased after the clustering-plus-resampling step. If niche weights or proposal parameters are chosen by an evolutionary heuristic rather than by a single global minimization of the KL divergence to the optimal importance density, the resulting mixture is no longer the solution of the original cross-entropy problem and a non-zero bias can appear; a short proof or explicit statement that the expectation is unchanged is required.

    Authors: We agree that an explicit statement is required. The niching operators (clustering and per-niche resampling) are applied once to the initial samples to construct a fixed mixture proposal density; the final importance-sampling estimator is then formed with respect to this fixed proposal. Because the importance weights are computed using the constructed proposal after niching, the estimator remains exactly unbiased for any proposal density. We will insert a short proof in the revised §3.2 showing that the expectation is unchanged by the deterministic post-processing step. revision: yes

  2. Referee: [§4] the reported failure-probability estimates for the multi-modal test cases lack accompanying standard-error or effective-sample-size diagnostics that would allow the reader to verify that the observed stability is not purchased at the cost of increased variance. Adding these quantities (or at least the coefficient of variation) for both the proposed method and the baseline IS/CE estimators would strengthen the empirical claim.

    Authors: We accept this suggestion. The revised §4 will report, for every method and every test case, the estimated coefficient of variation together with the effective sample size computed from the importance weights. These quantities will be added to the existing tables and discussed in the text so that readers can directly compare variance properties. revision: yes

  3. Referee: [§2.3] the paper should clarify whether the per-niche CE problems are solved independently or whether a global normalization of the mixture weights is performed after niching. The current description leaves open the possibility that the final proposal is no longer the minimizer of the single global KL objective, which would invalidate the usual optimality guarantees.

    Authors: We will revise the description in §2.3 to state explicitly that the relative cross-entropy problems are solved independently per niche and that the resulting component densities are then combined with globally normalized mixture weights. We will also add a remark noting that the resulting proposal is a practical approximation rather than the exact global minimizer of the single KL objective; the manuscript’s claims concern robustness on multi-modal problems rather than global optimality. revision: partial

standing simulated objections not resolved
  • A formal proof that the niching-derived mixture proposal remains the exact minimizer of the global cross-entropy objective; such a proof would require additional theoretical analysis that is outside the scope of the present work.

Circularity Check

0 steps flagged

No circularity: novel combination tested on independent examples

full rationale

The paper introduces niching importance sampling as a new framework that merges niching from evolutionary optimization with importance sampling and relative cross-entropy minimization. Its central claims rest on empirical performance across numerical test cases rather than any derivation that reduces a result to a fitted parameter, self-defined quantity, or load-bearing self-citation. No equation or step equates an output to its input by construction; the method is presented as a proposal whose bias and variance properties are assessed externally via simulation benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract alone provides insufficient detail to enumerate specific free parameters, axioms, or invented entities; the method is described only at the level of combining existing concepts from two fields.

pith-pipeline@v0.9.0 · 5375 in / 1123 out tokens · 25980 ms · 2026-05-10T17:52:26.784977+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

46 extracted references · 38 canonical work pages

  1. [2]

    H. AM, N. Lind, An Exact and Invariant First Order Reliability Format, Journal of Engineering Me- chanics 100 (Jan. 1974)

    1974

  2. [3]

    Backpropagation through time and the brain.Current Opinion in Neurobiology, 55:82–89, 2019

    B. Keshtegar, S. Chakraborty, A hybrid self-adaptive conjugate first order reliability method for robust structural reliability analysis, Applied Mathematical Modelling 53 (2018) 319–332.doi:10.1016/j. apm.2017.09.017

    work page doi:10.1016/j 2018

  3. [4]

    Der Kiureghian, H.-Z

    A. Der Kiureghian, H.-Z. Lin, S.-J. Hwang, Second-Order Reliability Approximations, Journal of En- gineering Mechanics 113 (8) (1987) 1208–1225.doi:10.1061/(ASCE)0733-9399(1987)113:8(1208)

    work page doi:10.1061/(asce)0733-9399(1987)113:8(1208 1987

  4. [5]

    A new direct second-order re- liability analysis method.Applied Mathematical Modelling, 55:68–80, 2018

    X. Huang, Y. Li, Y. Zhang, X. Zhang, A new direct second-order reliability analysis method, Applied Mathematical Modelling 55 (2018) 68–80.doi:10.1016/j.apm.2017.10.026

    work page doi:10.1016/j.apm.2017.10.026 2018

  5. [6]

    C. G. Bucher, U. Bourgund, A fast and efficient response surface approach for structural reliability problems, Structural Safety 7 (1) (1990) 57–66.doi:10.1016/0167-4730(90)90012-E

    work page doi:10.1016/0167-4730(90)90012-e 1990

  6. [7]

    Kaymaz, Application of Kriging method to structural reliability, Structural Safety 27 (2005) 133–151

    I. Kaymaz, Application of Kriging method to structural reliability, Structural Safety 27 (2005) 133–151. doi:10.1016/j.strusafe.2004.09.001

    work page doi:10.1016/j.strusafe.2004.09.001 2005

  7. [8]

    J. M. Bourinet, Rare-event probability estimation with adaptive support vector regression surrogates, Reliability Engineering & System Safety 150 (2016) 210–221.doi:10.1016/j.ress.2016.01.023

    work page doi:10.1016/j.ress.2016.01.023 2016

  8. [9]

    J. B. Cardoso, J. R. de Almeida, J. M. Dias, P. G. Coelho, Structural reliability analysis using Monte Carlo simulation and neural networks, Advances in Engineering Software 39 (6) (2008) 505–513.doi: 10.1016/j.advengsoft.2007.03.015

    work page doi:10.1016/j.advengsoft.2007.03.015 2008

  9. [10]

    G. Schueller, Efficient Monte Carlo simulation procedures in structural uncertainty and reliability anal- ysis - recent advances, Structural Engineering and Mechanics 32 (May 2009).doi:10.12989/sem. 2009.32.1.001

    work page doi:10.12989/sem 2009

  10. [11]

    Tabandeh, G

    A. Tabandeh, G. Jia, P. Gardoni, A review and assessment of importance sampling methods for relia- bility analysis, Structural Safety 97 (2022) 102216.doi:10.1016/j.strusafe.2022.102216

    work page doi:10.1016/j.strusafe.2022.102216 2022

  11. [12]

    Melchers, Importance sampling in structural systems, Structural Safety 6 (1) (1989) 3–10.doi: 10.1016/0167-4730(89)90003-9

    R. Melchers, Importance sampling in structural systems, Structural Safety 6 (1) (1989) 3–10.doi: 10.1016/0167-4730(89)90003-9

    work page doi:10.1016/0167-4730(89)90003-9 1989

  12. [13]

    C. G. Bucher, Adaptive sampling — an iterative fast Monte Carlo procedure, Structural Safety 5 (2) (1988) 119–126.doi:10.1016/0167-4730(88)90020-3

    work page doi:10.1016/0167-4730(88)90020-3 1988

  13. [14]

    Melchers, Search-based importance sampling, Structural Safety 9 (2) (1990) 117–128.doi:10.1016/ 0167-4730(90)90003-8

    R. Melchers, Search-based importance sampling, Structural Safety 9 (2) (1990) 117–128.doi:10.1016/ 0167-4730(90)90003-8

    1990

  14. [15]

    S. K. Au, J. L. Beck, Estimation of small failure probabilities in high dimensions by subset simulation, Probabilistic Engineering Mechanics 16 (4) (2001) 263–277.doi:10.1016/S0266-8920(01)00019-4

    work page doi:10.1016/s0266-8920(01)00019-4 2001

  15. [16]

    Papaioannou, C

    I. Papaioannou, C. Papadimitriou, D. Straub, Sequential importance sampling for structural reliability analysis, Structural Safety 62 (2016) 66–75.doi:10.1016/j.strusafe.2016.06.002

    work page doi:10.1016/j.strusafe.2016.06.002 2016

  16. [17]

    Kriaa, L

    I. Papaioannou, S. Geyer, D. Straub, Improved cross entropy-based importance sampling with a flexible mixture model, Reliability Engineering & System Safety 191 (2019) 106564.doi:10.1016/j.ress. 2019.106564. 28

    work page doi:10.1016/j.ress 2019

  17. [18]

    K. Breitung, The geometry of limit state function graphs and subset simulation: Counterexamples, Reliability Engineering & System Safety 182 (2019) 98–106.doi:10.1016/j.ress.2018.10.008

    work page doi:10.1016/j.ress.2018.10.008 2019

  18. [19]

    Breitung, The return of the design points, Reliability Engineering & System Safety 247 (2024) 110103.doi:10.1016/j.ress.2024.110103

    K. Breitung, The return of the design points, Reliability Engineering & System Safety 247 (2024) 110103.doi:10.1016/j.ress.2024.110103

    work page doi:10.1016/j.ress.2024.110103 2024

  19. [20]

    P. R. Wagner, S. Marelli, I. Papaioannou, D. Straub, B. Sudret, Rare event estimation using stochastic spectral embedding, Structural Safety 96 (2022) 102179.doi:10.1016/j.strusafe.2021.102179

    work page doi:10.1016/j.strusafe.2021.102179 2022

  20. [21]

    Rashki, SESC: A new subset simulation method for rare-events estimation, Mechanical Systems and Signal Processing 150 (2021) 107139.doi:10.1016/j.ymssp.2020.107139

    M. Rashki, SESC: A new subset simulation method for rare-events estimation, Mechanical Systems and Signal Processing 150 (2021) 107139.doi:10.1016/j.ymssp.2020.107139

    work page doi:10.1016/j.ymssp.2020.107139 2021

  21. [22]

    T. H. W. Bäck, A. V. Kononova, B. van Stein, H. Wang, K. A. Antonov, R. T. Kalkreuth, J. de Nobel, D. Vermetten, R. de Winter, F. Ye, Evolutionary Algorithms for Parameter Optimization—Thirty Years Later, Evolutionary Computation 31 (2) (2023) 81–122.doi:10.1162/evco_a_00325

    work page doi:10.1162/evco_a_00325 2023

  22. [23]

    X. Li, M. G. Epitropakis, K. Deb, A. Engelbrecht, Seeking Multiple Solutions: An Updated Survey on Niching Methods and Their Applications, IEEE Transactions on Evolutionary Computation 21 (4) (2017) 518–538.doi:10.1109/TEVC.2016.2638437

    work page doi:10.1109/tevc.2016.2638437 2017

  23. [24]

    Der Kiureghian, T

    A. Der Kiureghian, T. Dakessian, Multiple design points in first and second-order reliability, Structural Safety 20 (1) (1998) 37–49.doi:10.1016/S0167-4730(97)00026-X

    work page doi:10.1016/s0167-4730(97)00026-x 1998

  24. [25]

    H. J. Kinnear, F. DiazDelaO, Niching subset simulation, Probabilistic Engineering Mechanics 79 (2025) 103729.doi:10.1016/j.probengmech.2025.103729

    work page doi:10.1016/j.probengmech.2025.103729 2025

  25. [26]

    and Leibler, R

    M. Rosenblatt, Remarks on a Multivariate Transformation, The Annals of Mathematical Statistics 23 (3) (1952) 470–472.doi:10.1214/aoms/1177729394

    work page doi:10.1214/aoms/1177729394 1952

  26. [27]

    Nataf, Determination des distributions dont les marges sont donnees., C

    A. Nataf, Determination des distributions dont les marges sont donnees., C. R. Acad. Sci. Paris (225) (1962) 42–43

    1962

  27. [28]

    Kurtz, J

    N. Kurtz, J. Song, Cross-entropy-based adaptive importance sampling using Gaussian mixture, Struc- tural Safety 42 (2013) 35–44.doi:10.1016/j.strusafe.2013.01.006

    work page doi:10.1016/j.strusafe.2013.01.006 2013

  28. [29]

    Geyer, I

    S. Geyer, I. Papaioannou, D. Straub, Cross entropy-based importance sampling using Gaussian densities revisited, Structural Safety 76 (2019) 15–27.doi:10.1016/j.strusafe.2018.07.001

    work page doi:10.1016/j.strusafe.2018.07.001 2019

  29. [30]

    Z. Wang, J. Song, Cross-entropy-based adaptive importance sampling using von Mises-Fisher mixture for high dimensional reliability analysis, Structural Safety 59 (2016) 42–52.doi:10.1016/j.strusafe. 2015.11.002

    work page doi:10.1016/j.strusafe 2016

  30. [31]

    M. B. Mehni, M. B. Mehni, Reliability analysis with cross-entropy based adaptive Markov chain im- portance sampling and control variates, Reliability Engineering & System Safety 231 (2023) 109014. doi:10.1016/j.ress.2022.109014

    work page doi:10.1016/j.ress.2022.109014 2023

  31. [32]

    Klarbring, B

    A. Klarbring, B. Torstenfelt, Dynamical systems and topology optimization, Structural and Multidis- ciplinary Optimization 42 (2) (2010) 179–192.doi:10.1007/s00158-010-0479-9

    work page doi:10.1007/s00158-010-0479-9 2010

  32. [33]

    Z. Wang, M. Broccardo, J. Song, Hamiltonian Monte Carlo methods for Subset Simulation in reliability analysis, Structural Safety 76 (2019) 51–67.doi:10.1016/j.strusafe.2018.05.005

    work page doi:10.1016/j.strusafe.2018.05.005 2019

  33. [34]

    Friedli, D

    L. Friedli, D. Ginsbourger, A. Doucet, N. Linde, An energy-based model approach to rare event prob- ability estimation, SIAM/ASA Journal on Uncertainty Quantification 13 (2025) 400–424

    2025

  34. [35]

    Klawonn, F

    F. Klawonn, F. Höppner, B. Jayaram, What are Clusters in High Dimensions and are they Difficult to Find?, in: F. Masulli, A. Petrosino, S. Rovetta (Eds.), Clustering High–Dimensional Data, Springer, Berlin, Heidelberg, 2015, pp. 14–33.doi:10.1007/978-3-662-48577-4_2. 29

    work page doi:10.1007/978-3-662-48577-4_2 2015

  35. [36]

    X. Li, A. P. Engelbrecht, M. G. Epitropakis, Benchmark functions for the cec 2013 special session and competition on niching methods for multimodal function optimization, in: Proceedings of the IEEE Congress on Evolutionary Computation (CEC 2013), IEEE, 2013, pp. 1–10

    2013

  36. [37]

    Ursem, Multinational evolutionary algorithms, in: Proceedings of the 1999 Congress on Evolutionary Computation-CEC99 (Cat

    R. Ursem, Multinational evolutionary algorithms, in: Proceedings of the 1999 Congress on Evolutionary Computation-CEC99 (Cat. No. 99TH8406), IEEE, Washington, DC, USA, 1999, pp. 1633–1640.doi: 10.1109/CEC.1999.785470

    work page doi:10.1109/cec.1999.785470 1999

  37. [38]

    G. Jia, A. Tabandeh, P. Gardoni, A density extrapolation approach to estimate failure probabilities, Structural Safety 93 (2021) 102128.doi:10.1016/j.strusafe.2021.102128

    work page doi:10.1016/j.strusafe.2021.102128 2021

  38. [39]

    M. A. Valdebenito, H. J. Pradlwarter, G. I. Schuëller, The role of the design point for calculating failure probabilities in view of dimensionality and structural nonlinearities, Structural Safety 32 (2) (2010) 101–111.doi:10.1016/j.strusafe.2009.08.004

    work page doi:10.1016/j.strusafe.2009.08.004 2010

  39. [40]

    Katafygiotis, K

    L. Katafygiotis, K. Zuev, Geometric insight into the challenges of solving high-dimensional reliabil- ity problems, Probabilistic Engineering Mechanics 23 (2008) 208–218.doi:10.1016/j.probengmech. 2007.12.026

    work page doi:10.1016/j.probengmech 2008

  40. [41]

    A.Abdi, M.Kaveh, PerformancecomparisonofthreedifferentestimatorsfortheNakagamimparameter using Monte Carlo simulation, IEEE Communications Letters 4 (4) (2000) 119–121.doi:10.1109/ 4234.841316

    work page arXiv 2000

  41. [42]

    S. Sra, A short note on parameter approximation for von Mises-Fisher distributions: And a fast implementation of I s (x), Computational Statistics 27 (1) (2012) 177–190.doi:10.1007/ s00180-011-0232-x

    2012

  42. [43]

    Ester, H.-P

    M. Ester, H.-P. Kriegel, J. Sander, X. Xu, A density-based algorithm for discovering clusters in large spatial databases with noise, in: Proceedings of the Second International Conference on Knowledge Discovery and Data Mining, KDD’96, AAAI Press, Portland, Oregon, 1996, pp. 226–231

    1996

  43. [44]

    J. MacQueen, Some methods for classification and analysis of multivariate observations, in: Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, Volume 1: Statistics, Vol. 5.1, University of California Press, 1967, pp. 281–298

    1967

  44. [45]

    Comput- ers &; Security134, 103448 (2023)

    A. Sharma, C. S. Manohar, Modified replica exchange-based MCMC algorithm for estimation of struc- tural reliability based on particle splitting method, Probabilistic Engineering Mechanics 72 (2023) 103448.doi:10.1016/j.probengmech.2023.103448

    work page doi:10.1016/j.probengmech.2023.103448 2023

  45. [46]

    Rashki, M

    M. Rashki, M. Miri, M. A. Moghaddam, A simulation-based method for reliability based design op- timization problems with highly nonlinear constraints, Automation in Construction 47 (2014) 24–36. doi:10.1016/j.autcon.2014.07.004

    work page doi:10.1016/j.autcon.2014.07.004 2014

  46. [47]

    El Masri, J

    M. El Masri, J. Morio, F. Simatos, Improvement of the cross-entropy method in high dimension for failure probability estimation through a one-dimensional projection without gradient estimation, Reli- ability Engineering & System Safety 216 (2021) 107991.doi:10.1016/j.ress.2021.107991. 30

    work page doi:10.1016/j.ress.2021.107991 2021