Christian Léger

Bootstrap technology and applications

Bootstrap resampling methods have emerged as powerful tools for constructing inferential procedures in modern statistical data analysis. Although these methods depend on the availability of fast, inexpensive computing, they offer the potential for highly accurate methods of inference. Moreover, they can even eliminate the need to impose a convenient statistical model that does not have a strong scientific basis. In this article, we review some bootstrap methods, emphasizing applications through illustrations with some real data. Special attention is given to regression, problems with dependent data, and choosing tuning parameters for optimal performance.

Bandwidth selection for kernel distribution function estimation

Abstract Leave-one-out cross-validation is a popular and readily implemented heuristic for bandwidth selection in nonparametric smoothing problems. In this note we elucidate the role of leave-one-out selection criteria by discussing a criterion introduced by Sarda (J. Statist. Plann. Inference 35 (1993) 65–75) for bandwidth selection for kernel distribution function estimators (KDFEs). We show that for this problem, use of the leave-one-out KDFE in the selection procedure is asymptotically equivalent to leaving none out. This contrasts with kernel density estimation, where use of the leave-one-out density estimator in the selection procedure is critical. Unfortunately, simulations show that neither method works in practice, even for samples of size as large as 1000. In fact, we show that for any fixed bandwidth, the expected value of the derivative of the leave-none-out criterion is asymptotically positive. This result and our simulations suggest that the criteria are increasing and that for sufficiently large samples (e.g., n = 100), the smallest available bandwidth will always be selected, thus contradicting the optimality result of Sarda for this estimator. As an alternative to minimizing a selection criterion, we propose a plug-in estimator of the asymptotically optimal bandwidth. Simulations suggest that the plug-in is a good estimator of the asymptotically optimal bandwidth even for samples as small as 10 observations and is not too far from the finite sample bandwidth.

Bootstrap choice of tuning parameters

AbstractConsider the problem of estimating θ=θ(P) based on dataxn from an unknown distributionP. Given a family of estimatorsTn, β of θ(P), the goal is to choose β among β∈I so that the resulting estimator is as good as possible. Typically, β can be regarded as a tuning or smoothing parameter, and proper choice of β is essential for good performance ofTn, β. In this paper, we discuss the theory of β being chosen by the bootstrap. Specifically, the bootstrap estimate of β,

Assessing influence in variable selection problems

\hat \beta _n

Nonparametric age replacement: bootstrap confidence intervals for the optimal cost

, is chosen to minimize an empirical bootstrap estimate of risk. A general theory is presented to establish the consistency and weak convergence properties of these estimators. Confidence intervals for θ(P) based on

Bootstrapping regression models with BLUS residuals

T_{n,\hat \beta _n }