Abstract
In this note, we consider the problem of estimating the smaller of two ordered means. Such problems frequently arise in applications where, for example, aggregated data are observed. In order to combine information from direct and indirect observations, we use the Stein-type truncated estimator. We show that it dominates the direct estimator for distributions with log-concave or log-convex densities.
References
Bagnoli M, Bergstrom T (2005) Log-concave probability and its applications. Econ Theor 26:445–469
Bhattacharya CG (1984) Two inequalities with an application. Ann Inst Stat Math 36:129–134
Blumenthal S, Cohen A (1968a) Estimation of the larger translation parameter. Ann Math Stat 39:502–516
Blumenthal S, Cohen A (1968b) Estimation of two ordered translation parameters. Ann Math Stat 39:517–530
Chang Y-T, Shinozaki N (2006) Estimation of ordered means of two Poisson distributions. Commun Stat Theory Methods 35:1993–2003
Cohen A, Sackrowitz HB (1970) Estimation of the last mean of a monotone sequence. Ann Math Stat 41:2021–2034
Garg N, Misra N (2021) Componentwise equivariant estimation of order restricted location and scale parameters in bivariate models: a unified study. Preprint at http://arxiv.org/abs/2109.14997
Hudson HM (1978) A natural identity for exponential families with applications in multiparameter estimation. Ann Stat 6:473–484
Hwang JT, Peddada SD (1994) Confidence interval estimation subject to order restrictions. Ann Stat 22:67–93
Jena AK, Tripathy MR (2019) Estimating ordered quantiles of two exponential populations with a common minimum guarantee time. Commun Stat Theory Methods 48:3570–3585
Katz MW (1963) Estimating ordered probabilities. Ann Math Stat 34:967–972
Kelly RE (1989) Stochastic reduction of loss in estimating normal means by isotonic regression. Ann Stat 17:937–940
Kubokawa T, Saleh AKMDE (1994) Estimation of location and scale parameters under order restrictions. J Stat Res 28:41–51
Kumar S, Sharma D (1988) Simultaneous estimation of ordered parameters. Commun Stat Theory Methods 17:4315–4336
Kumar S, Sharma D (1989) On the pitman estimator op ordered normal means. Commun Stat Theory Methods 18:4163–4175
Kumar S, Sharma D (1993) Minimaxity of the Pitman estimator of ordered normal means when variances are unequal. J Indian Soc Agric Stat 45:230–234
Kumar S, Kumar A, Tripathi YM (2005a) A Note on the Pitman estimator of ordered normal means when the variances are unequal. Commun Stat Theory Methods 34:2115–2122
Kumar S, Tripathi YM, Misra N (2005b) James–Stein type estimators for ordered normal means. J Stat Comput Simul 75:501–511
Kushary D, Cohen A (1991) Estimation of ordered Poisson parameters. Sankhyā Indian J Stat Ser A 53:334–356
Lee CIC (1981) The quadratic loss of isotonic regression under normality. Ann Stat 9:686–688
Lee CIC (1988) Quadratic loss of order restricted estimators for treatment means with a control. Ann Stat 16:751–758
Misra N, Choudhary PK, Dhariyal ID, Kundu D (2002) Smooth estimators for estimating order restricted scale parameters of two gamma distributions. Metrika 56:143–161
Misra N, Iyer SK, Singh H (2004) The linex risk of maximum likelihood estimators of parameters of normal populations having order restricted means. Sankhyā Indian J Stat 66:652–677
Patra LK, Kumar S (2017) Estimating ordered means of a bivariate normal distribution. Am J Math Manag Sci 36:118–136
Patra LK, Kumar S, Petropoulos C (2021) Componentwise estimation of ordered scale parameters of two exponential distributions under a general class of loss function. Statistics 55:595–617
Polson NG, Scott JG, Windle J (2013) Bayesian inference for logistic models using pólya-gamma latent variables. J Am Stat Assoc 108:1339–1349
Shonkwiler JS (2016) Variance of the truncated negative binomial distribution. J Econometr 195:209–210
Singh H, Misra N, Li S (2005) Estimation of order restricted concentration parameters of von Mises distributions. Commun Stat Simul Comput 34:21–40
Stein C (1964) Inadmissibility of the usual estimator for the variance of a normal distribution with unknown mean. Ann Inst Stat Math 16:155–160
van Eeden C (2006) Restricted parameter space problems—admissibility and minimaxity properties. Lecture notes in statistics. Springer, New York
Vijayasree G, Singh H (1991) Simultaneous estimation of two ordered exponential parameters. Commun Stat Theory Methods 20:2559–2576
Vijayasree G, Misra N, Singh H (1995) Componentwise estimation of ordered parameters of \(k(\ge 2)\) exponential populations. Ann Inst Stat Math 47:287–307
Acknowledgements
We would like to thank the editor, the associate editor, and the two reviewers for many valuable comments and helpful suggestions which improved the paper. Research of the authors was supported in part by Grant-in-Aid for Scientific Research (22K20132, 20J10427, 19K11852, 18K11188) from Japan Society for the Promotion of Science.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
1.1 Proofs
Here we prove Propositions 2.2, 3.1, 3.2 and Lemma 2.2.
Proof of Proposition 2.2. We have
Now, by assumption,
Therefore, \(z f(z / (1 + v)) / f(z)\) is a nondecreasing function of \(z > 0\), which implies that \(\int _{0}^{x} z f(z / (1 + v)) dz / \int _{0}^{x} f(z) dz\) is a nondecreasing function of \(x > 0\). Thus, by the covariance inequality, the right-hand side of (5.1) is nonnegative. The result follows. \(\square \)
Proof of Lemma 2.2. We prove parts (i) and (ii) based on the proofs of Lemmas 3 and 4 of Bagnoli and Bergstrom (2005), which are for the continuous case. Without loss of generality, we can assume that \(k = 1\) and \(x_2 = x_1 + 1\). Let \(F(- 1) = f(- 1) = 0\).
For part (i), note that \(f(x - 1) / f(x) \ge f(z - 1) / f(z)\) for all \(x, z \in \mathbb {N} _0\) with \(x \ge z\) by the log-concavity of f. Then, for all \(x \in \mathbb {N} _0\),
which implies that
Therefore,
For part (ii), note that \(f(x) / f(x + 1) \ge f(z) / f(z + 1)\) for all \(x, z \in \mathbb {N} _0\) with \(x \le z\) by the log-convexity of f. Then, for all \(x \in \mathbb {N} _0\),
where the last equality follows from the assumption that \(\lim _{z \rightarrow \infty } f(z) = 0\), and thus we have
Hence,
For part (iii), let \(W = X + Z\) and \(T = X - Z\). Let \(S = 1(W \in 2 \mathbb {N} _0 + 1)\). Suppose first that \(\{ \partial / ( \partial x) \} \{ \log f(x) \} \) is a convex function of \(x \in [0, \infty )\) and that \(E[ W | S = 0 ] \le E[ W | S = 1 ]\). Then
and
Note that for \(w \in \mathbb {N} _0\),
where
is a nondecreasing function of \(w \in 2 \mathbb {N} _0\) by assumption. Then it follows from Theorem 2.1 of Bhattacharya (1984) that
since E[W|S] and \(E[ (1 - S) \rho (W) + S | S]\) are nondecreasing functions of S by assumption and since W and \((1 - S) \rho (W) + S\) are nondecreasing functions of W. Next, suppose instead that f is log-convex. Then
where the first inequality follows from the covariance inequality since \(f(z + T) / f(z)\) is a nondecreasing function of \(z \in \mathbb {N} _0\) by assumption. Thus,
This completes the proof. \(\square \)
Proof of Proposition 3.1. Let \(W = X + Y\) and \(R = X / W\). Let \({\Delta }= E [ ( \max \{ Y, (X + Y) / 2 \} - {\alpha }_2 )^2 ] - E[ (Y - {\alpha }_2 )^2 ]\). Note that \(W \sim \mathrm{{Ga}} ( {\alpha }_1 + {\alpha }_2 , 1)\) and \(R \sim \mathrm{{Beta}} ( {\alpha }_1 , {\alpha }_2 )\) and these are mutually independent.
For part (i), suppose that \({\alpha }_1 = {\alpha }_2\). Then it can be seen that
as \({\alpha }_2 \rightarrow 0\).
For part (ii), let \(\phi (R) = \max \{ Y, (X + Y) / 2 \} / Y = \max \{ 1, 1 / \{ 2 (1 - R) \} \} \). Then
Since \(\{ \phi (R) + 1 \} (1 - R) \le 1\) when \(\phi (R) - 1 \ne 0\) and since \({\alpha }_1 + {\alpha }_2 + 1 < 2 {\alpha }_2\) by assumption, it follows that \({\Delta }< 0\). \(\square \)
Proof of Proposition 3.2. Let \(W = X + Y\) and \(R = X / W\). For part (i), let \(\phi _1 (R) = \min \{ X, (X + Y) / 2 \} / X = \min \{ 1, 1 / (2 R) \} \). Then we have
which is negative since \(( {\alpha }_1 + {\alpha }_2 ) / {\alpha }_1 \ge 2\) by assumption and since \(2 R \{ 1 - \phi _1 (R) \} + \log \phi _1 (R) > 0\) when \(\phi _1 (R) \ne 1\). For part (ii), let \(\phi _2 (R) = \max \{ 1, 1 / \{ 2 (1 - R) \} \} \). Then
which is negative since \(( {\alpha }_1 + {\alpha }_2 ) / {\alpha }_1 \le 2\) and since \(2 (1 - R) \{ \phi _2 (R) - 1 \} - \log \phi _2 (R) < 0\) when \(\phi _2 (R) \ne 1\). \(\square \)
1.2 Details of the approximation algorithm used in Sect. 4.5
First, the density proportional to
is approximated by \(\mathrm{{Ga}} (1 / 2, \log 2)\) because for all \(k \in \mathbb {N} _0\) and all \(\mu \in (0, \infty )\),
provided that \(\mu > 1\), and because
for all \(\mu \in (0, \infty )\) (see Hudson (1978) for the equality). Next, note that for all \(K \in \mathbb {N}\) and all \(\mu \in (0, \infty )\),
by Jensen’s inequality, where \({\widetilde{X}}\sim \mathrm{{NB}} ( \mu , 1 / 2)\) and where \(E[ {\widetilde{X}}| {\widetilde{X}}> K ]\) is expressible in closed form (see Shonkwiler (2016)). Then, based on these inequalities, the acceptance ratio can be bounded below and above with arbitrary accuracy in MH steps (such an approach is used, for example, by Polson et al. (2013)).
Rights and permissions
About this article
Cite this article
Hamura, Y., Kubokawa, T. Robustness of a truncated estimator for the smaller of two ordered means. Stat Papers 64, 2225–2244 (2023). https://doi.org/10.1007/s00362-022-01371-3
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00362-022-01371-3