A General Framework for Bayes Structured Linear Models
نویسندگان
چکیده
High dimensional statistics deals with the challenge of extracting structured information from complex model settings. Compared with the growing number of frequentist methodologies, there are rather few theoretically optimal Bayes methods that can deal with very general high dimensional models. In contrast, Bayes methods have been extensively studied in various nonparametric settings and rate optimal posterior contraction results have been established. This paper provides a unified approach to both Bayes high dimensional statistics and Bayes nonparametrics in a general framework of structured linear models. With the proposed two-step model selection prior, we prove a general theorem of posterior contraction under an abstract setting. The main theorem can be used to derive new results on optimal posterior contraction under many complex model settings including stochastic block model, graphon estimation and dictionary learning. It can also be used to re-derive optimal posterior contraction for problems such as sparse linear regression and nonparametric aggregation, which improve upon previous Bayes results for these problems. The key of the success lies in the proposed two-step prior distribution. The prior on the parameters is an elliptical Laplace distribution that is capable to model signals with large magnitude, and the prior on the models involves an important correction factor that compensates the effect of the normalizing constant of the elliptical Laplace distribution.
منابع مشابه
The Family of Scale-Mixture of Skew-Normal Distributions and Its Application in Bayesian Nonlinear Regression Models
In previous studies on fitting non-linear regression models with the symmetric structure the normality is usually assumed in the analysis of data. This choice may be inappropriate when the distribution of residual terms is asymmetric. Recently, the family of scale-mixture of skew-normal distributions is the main concern of many researchers. This family includes several skewed and heavy-tailed d...
متن کاملPAC-Bayes Generalization Bounds for Randomized Structured Prediction
We present a new PAC-Bayes generalization bound for structured prediction that is applicable to perturbation-based probabilistic models. Our analysis explores the relationship between perturbation-based modeling and the PAC-Bayes framework, and connects to recently introduced generalization bounds for structured prediction. We obtain the first PAC-Bayes bounds that guarantee better generalizati...
متن کاملNon-linear Bayesian prediction of generalized order statistics for liftime models
In this paper, we obtain Bayesian prediction intervals as well as Bayes predictive estimators under square error loss for generalized order statistics when the distribution of the underlying population belongs to a family which includes several important distributions.
متن کاملStructured Prediction: From Gaussian Perturbations to Linear-Time Principled Algorithms
Margin-based structured prediction commonly uses amaximum loss over all possible structured outputs [1, 4,14]. In natural language processing, recent work [15, 16]has proposed the use of the maximum loss over randomstructured outputs sampled independently from someproposal distribution. This method is linear-time in thenumber of random structured outputs and trivially pa...
متن کاملBayesian model selection in complex linear systems, as illustrated in genetic association studies.
Motivated by examples from genetic association studies, this article considers the model selection problem in a general complex linear model system and in a Bayesian framework. We discuss formulating model selection problems and incorporating context-dependent a priori information through different levels of prior specifications. We also derive analytic Bayes factors and their approximations to...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2015