arXiv subjects

Luis R. Pericchi

Publications and source records attributed to Luis R. Pericchi.

3 recordsLinked to original sources

arXiv2025

Limiting the Shrinkage for the Exceptional by Objective Robust Bayesian Analysis: the `Clemente Problem'

In `borrowing strength' an important problem of Statistics is to treat exceptional cases in a fundamentally different. This is what has been coined as `the Clemente problem' in honor of R. Clemente (Efron 2010). In this article, we propose to use robust penalties, in the form of losses that penalize more severely huge errors, or (equivalently) priors of heavy tails which make more probable the exceptional. Using heavy tailed priors, we can reproduce in a Bayesian way Efron and Morris `limited translated estimators' (with Double Exponential Priors) and `discarding priors estimators' (with Cauchy like priors), which discard the prior in the presence of conflict. Both Empirical Bayes and Full Bayes approaches are able to alleviate the Clemente Problem and furthermore beat the James-Stein estimator in terms of smaller square errors, for sensible Robust Bayes priors. We model in parallel Empirical Bayes and Fully Bayesian hierarchical models, illustrating that the differences among sensible versions of both are relatively small, as compared with the effect due to the robust assumptions.

stat.ME↗

arXiv2017

Converting P-Values in Adaptive Robust Lower Bounds of Posterior Probabilities to increase the reproducible Scientific "Findings"

We put forward a novel calibration of p values, the "Adaptive Robust Lower Bound" (ARLB) which maps p values into approximations of posterior probabilities taking into account the effect of sample sizes. We build on the Robust Lower Bound proposed by Sellke, Bayarri and Berger (2001), but we incorporate a simple power of the sample size to make it adaptive to different amounts of data. We present several illustrations from where it is apparent that the ARLB closely approximates exact Bayes Factors. In particular, it has the same asymptotics as posterior probabilities but avoiding the problems of "Bayesian Information Criterion" (BIC) for small samples relative to the number of parameters. We prove that the ARLB is consistent as the sample size grows, and that it is information consistent (Berger and Pericchi, 2001) for the canonical Normal case, but with methods that are keen to be generalized. So ARLB also avoids the problems of certain conjugate priors as g-priors. In summary, this is a novel criterion easy to apply, as it only requires a real p value, a sample size and parameter dimensionality. This method is intended to aid the practitioners, who are increasingly aware of the lack of reproducibility of traditional hypothesis testing "findings" but at the same time, lack of concrete simple alternatives. Here is one.

stat.ME↗

arXiv2004

Training samples in objective Bayesian model selection

Central to several objective approaches to Bayesian model selection is the use of training samples (subsets of the data), so as to allow utilization of improper objective priors. The most common prescription for choosing training samples is to choose them to be as small as possible, subject to yielding proper posteriors; these are called minimal training samples. When data can vary widely in terms of either information content or impact on the improper priors, use of minimal training samples can be inadequate. Important examples include certain cases of discrete data, the presence of censored observations, and certain situations involving linear models and explanatory variables. Such situations require more sophisticated methods of choosing training samples. A variety of such methods are developed in this paper, and successfully applied in challenging situations.

math.ST↗