Bounds on Sample Variation Measures Based on Majorization

Using majorization theory, upper and lower bounds are derived for different measures of variation as progressively more items of information are available about the sample data. As a convenient starting point, bounds are first established for a one-parameter family of variation measures, which is a generalized mean difference measure of which Gini's mean difference, the standard deviation, and the range are particular cases. While, as pointed out, some of the derived bounds are well known, others do not appear to have been published and are tighter than established bounds. Some 40 different bounds are derived, besides any number of bounds given for the generalized family of variation measures. A number of interesting inequalities are also derived on the basis of some of the bounds. While the bounds have been developed in terms of real-valued sample data generally, the paper concludes with a brief discussion of the bounds for categorical data when the sample data consists of frequencies (counts).     

基于分组数据的样本基尼系数范围估计

 由于很多收入抽样数据只是公布了相对简约的分组数据,如何依据信息不完整的分组数据估计样本基尼系数的范围是非常重要的。本文针对分组数据中各组收入的取值范围以及各组人均收入是否已知,从多个方面探讨了样本基尼系数的取值范围,并给出了相应的估算公式。最后,结合我国城乡居民收入的分组数据,实证分析了城乡收入基尼系数的范围。     

Lower confidence bounds for percentiles of weibull and birnbaum-saunders distributions

Approximate lower confidence bounds on percentiles of the Weibull and the Birnbaum-Saunders distributions are investigated. Asymptotic lower confidence bounds based on Bonferroni's inequality and the Fisher information are discussed, and parametric bootstrap methods to provide better bounds are considered. Since the standard percentile bootstrap method typically does not perform well for confidence bounds on quantiles, several other bootstrap procedures are studied via extensive computer simulations. Results of the simulations indicate that the bootstrap methods generally give sharper lower bounds than the Bonferroni bounds but with coverages still near the nominal confidence level. Two illustrative examples are also presented, one for tensile strength of carbon micro-composite specimens and the other for cycles-to-failure data.     

Cramér-Rao lower bounds for estimators based on censored data

The Cramér-Rao lower bounds for the variances of unbiased estimators based on censored data are given. Useful techniques of evaluation are then derived for these lower bounds. Examples are given to illustrate these techniques. Small-sample comparisons are made between the resulting lower bounds, the variances of the best linear unbiased estimators, and the variances of unbiased esti-mators which are based on the maximum likelihood estimators.     

Low dose risk estimation via simultaneous statistical inferences

Summary.  The paper develops and studies simultaneous confidence bounds that are useful for making low dose inferences in quantitative risk analysis. Application is intended for risk assessment studies where human, animal or ecological data are used to set safe low dose levels of a toxic agent, but where study information is limited to high dose levels of the agent. Methods are derived for estimating simultaneous, one-sided, upper confidence limits on risk for end points measured on a continuous scale. From the simultaneous confidence bounds, lower confidence limits on the dose that is associated with a particular risk (often referred to as a bench-mark dose ) are calculated. An important feature of the simultaneous construction is that any inferences that are based on inverting the simultaneous confidence bounds apply automatically to inverse bounds on the bench-mark dose.     

Lower bounds for confidence coefficients for confidence intervals for finite population quantiles

Assuming stratified simple random sampling, a confidence interval for a finite population quantile may be desired. Using a confidence interval with endpoints given by order statistics from the combined stratified sample, several procedures to obtain lower bounds (and approximations for the lower bounds) for the confidence coefficients are presented. The procedures differ with respect to the amount of prior information assumed about the var-iate values in the finite population, and the extent to which sample data is used to estimate the lower bounds.     

Post hoc false positive control for structured hypotheses

In a high-dimensional multiple testing framework, we present new confidence bounds on the false positives contained in subsets S of selected null hypotheses. These bounds are post hoc in the sense that the coverage probability holds simultaneously over all S, possibly chosen depending on the data. This article focuses on the common case of structured null hypotheses, for example, along a tree, a hierarchy, or geometrically (spatially or temporally). Following recent advances in post hoc inference, we build confidence bounds for some prespecified forest-structured subsets and deduce a bound for any subset S by interpolation. The proposed bounds are shown to improve substantially previous ones when the signal is locally structured. Our findings are supported both by theoretical results and numerical experiments. Moreover, our bounds can be obtained by an algorithm (with complexity bilinear in the sizes of the reference hierarchy and of the selected subset) that is implemented in the open-source R package sansSouci available from https://github.com/pneuvial/sanssouci , making our approach operational.     

Probability density estimation via an infinite Gaussian mixture model: application to statistical process monitoring

Summary.  The primary goal of multivariate statistical process performance monitoring is to identify deviations from normal operation within a manufacturing process. The basis of the monitoring schemes is historical data that have been collected when the process is running under normal operating conditions. These data are then used to establish confidence bounds to detect the onset of process deviations. In contrast with the traditional approaches that are based on the Gaussian assumption, this paper proposes the application of the infinite Gaussian mixture model (GMM) for the calculation of the confidence bounds, thereby relaxing the previous restrictive assumption. The infinite GMM is a special case of Dirichlet process mixtures and is introduced as the limit of the finite GMM, i.e. when the number of mixtures tends to ∞. On the basis of the estimation of the probability density function, via the infinite GMM, the confidence bounds are calculated by using the bootstrap algorithm. The methodology proposed is demonstrated through its application to a simulated continuous chemical process, and a batch semiconductor manufacturing process.     

Algorithms for Computing the Sample Variance: Analysis and Recommendations

The problem of computing the variance of a sample of N data points {x_i } may be difficult for certain data sets, particularly when N is large and the variance is small. We present a survey of possible algorithms and their round-off error bounds, including some new analysis for computations with shifted data. Experimental results confirm these bounds and illustrate the dangers of some algorithms. Specific recommendations are made as to which algorithm should be used in various contexts.     

Epidemics in heterogeneous communities: estimation of R0 and secure vaccination coverage

A stochastic multitype model for the spread of an infectious disease in a community of heterogeneous individuals is analysed. In particular, estimates of R ₀ (the basic reproduction number) and the critical vaccination coverage are derived, where estimation is based on final size data of an outbreak in the community. It is shown that these key parameters cannot be estimated consistently from data; only upper and lower bounds can be estimated. Confidence regions for the upper bounds are derived, thus giving conservative estimates of R ₀ and the fractions necessary to vaccinate.     

In Memoriam: Jack Carl Kiefer 1924–1981

This note provides a new explanation for Tukey's definition of “(inner) fences” for box-and-whiskers plots. The starting point is explicit bounds for the sample mean based only on the box plot. Starting from these bounds we define a dataset to contain outside values if at least one of the latter bounds is outside of the box. This leads to a new, yet simple definition of fences. They are symmetric around the box if, and only if, the median is in the middle of the box. In that case, the new definition coincides with Tukey's rule of 1.5 times the inter quartile range. To avoid instabilities for small (sub-) samples we propose to complement the box-and-whiskers plot of the original data with a box-and-whiskers plot of Walsh means.     

Confidence Bounds For The Binomial N Parameter: A Classical Approach

A method is described for finding the distribution of an estimator for the parameter N of the binomial distribution in order to construct confidence bounds for N. Using a partition generating algorithm, tables of lower confidence bounds are provided for a range of values, illustrating the nature of the problems involved in estimating the parameters. Examples are considered and the lower bounds compared with the convenient approximate bounds due to Hoel (1947), suggesting the approximate bounds are appropriate more generally then originally thought.     

Bounding sample size projections for the area under a ROC curve

Studies of diagnostic tests are often designed with the goal of estimating the area under the receiver operating characteristic curve (AUC) because the AUC is a natural summary of a test's overall diagnostic ability. However, sample size projections dealing with AUCs are very sensitive to assumptions about the variance of the empirical AUC estimator, which depends on two correlation parameters. While these correlation parameters can be estimated from the available data, in practice it is hard to find reliable estimates before the study is conducted. Here we derive achievable bounds on the projected sample size that are free of these two correlation parameters. The lower bound is the smallest sample size that would yield the desired level of precision for some model, while the upper bound is the smallest sample size that would yield the desired level of precision for all models. These bounds are important reference points when designing a single or multi-arm study; they are the absolute minimum and maximum sample size that would ever be required. When the study design includes multiple readers or interpreters of the test, we derive bounds pertaining to the average reader AUC and the ‘pooled’ or overall AUC for the population of readers. These upper bounds for multireader studies are not too conservative when several readers are involved.     

Some New Bounds on the Probability of Extinction of a Galton–Watson Process with Numerical Comparisons

Some new upper and lower bounds for the extinction probability of a Galton–Watson process are presented. They are very easy to compute and can be used even if the offspring distribution has infinite variance. These new bounds are numerically compared to previously discussed bounds. Some definite guidelines are given concerning when these new bounds are preferable. Some open problems are also discussed.     

A Note on Window Length Selection in Singular Spectrum Analysis

In singular spectrum analysis (SSA) window length is a critical tuning parameter that must be assigned by the practitioner. This paper provides a theoretical analysis of signal–noise separation and time series reconstruction in SSA that can serve as a guide to optimal window choice. We establish numerical bounds on the mean squared reconstruction error and present their almost sure limits under very general regularity conditions on the underlying data generating mechanism. We also provide asymptotic bounds for the mean squared separation error. Evidence obtained using simulation experiments and real data sets indicates that the theoretical properties are reflected in observed behaviour, even in relatively small samples, and the results indicate how, in practice, an optimal assignment for the window length can be made.     

Search With Learning for Differentiated Products: Evidence from E-Commerce

This article provides a method to estimate search costs in a differentiated product environment in which consumers are uncertain about the utility distribution. Consumers learn about the utility distribution by Bayesian updating their Dirichlet process prior beliefs. The model provides expressions for bounds on the search costs that can rationalize observed search and purchasing behavior. Using individual-specific data on web browsing and purchasing behavior for MP3 players sold online we show how to use these bounds to estimate search costs as well as the parameters of the utility distribution. Our estimates indicate that search costs are sizable. We show that ignoring consumer learning while searching can lead to severely biased search cost and elasticity estimates.     

Max-min multiple comparison procedure for comparing several dose levels with a zero dose control

The comparison of increasing doses of a compound to a zero dose control is of interest in medical and toxicological studies. Assume that the mean dose effects are non-decreasing among the non-zero doses of the compound. A simple procedure that modifies Dunnett's procedure is proposed to construct simultaneous confidence intervals for pairwise comparisons of each dose group with the zero dose control by utilizing the ordering of the means. The simultaneous lower bounds and upper bounds by the new procedure are monotone, which is not the case with Dunnett's procedure. This is useful to categorize dose levels. The expected gains of the new procedure over Dunnett's procedure are studied. The procedure is shown by real data to compare well with its predecessor.     

Confidence estimation via the parametric bootstrap in logistic joinpoint regression

We consider asymptotic properties of the maximum likelihood and related estimators in a clustered logistic joinpoint model with an unknown joinpoint. Sufficient conditions are given for the consistency of confidence bounds produced by the parametric bootstrap; one of the conditions required is that the true location of the joinpoint is not at one of the observation times. A simulation study is presented to illustrate the lack of consistency of the bootstrap confidence bounds when the joinpoint is an observation time. A removal algorithm is presented which corrects this problem, but at the price of an increased mean square error. Finally, the methods are applied to data on yearly cancer mortality in the US for individuals age 65 and over.     

Bounds on the regression coefficients when a covariate is categorized

This paper considers the problem of a linear regression model in which a continuous covariate is categorized. In such a case one can obtain bounds on the OLS-estimate which would be calculated if the complete information were available. These bounds give an indication of the information loss due to grouping and show what can be inferred from the data without additional assumptions.     

Robust bayesian analysis given bounds on the probability of a set

Suppose that just the lower and the upper bounds on the probability of a measurable subset K in the parameter space ω are a priori known. Instead of eliciting a unique prior probability measure, consider the class Γ of all the probability measures compatible with such bounds. Under mild regularity conditions about the likelihood function, both prior and posterior bounds on the expected value of any function of the unknown parameter ω are computed, as the prior measure varies in Γ. Such bounds are analysed according to the robust Bayesian viewpoint. Furthermore, lower and upper bounds on the Bayes factor are corisidered. Finally, the local sensitivity analysis is performed, considering the class Γ as a aeighbourhood of an elicited prior