Conditional inference in parametric models
نویسندگان
چکیده
This paper presents a new approach to conditional inference, based on the simulation of samples conditioned by a statistics of the data. Also an explicit expression for the approximation of the conditional likelihood of long runs of the sample given the observed statistics is provided. It is shown that when the conditioning statistics is sufficient for a given parameter, the approximating density is still invariant with respect to the parameter. A new Rao-Blackwellisation procedure is proposed and simulation shows that Lehmann Scheffé Theorem is valid for this approximation. Conditional inference for exponential families with nuisance parameter is also studied, leading to Monte carlo tests. Finally the estimation of the parameter of interest through conditional likelihood is considered. Comparison with the parametric bootstrap method is discussed.
منابع مشابه
Exact Statistical Inference for Some Parametric Nonhomogeneous Poisson Processes
Nonhomogeneous Poisson processes (NHPPs) are often used to model recurrent events, and there is thus a need to check model fit for such models. We study the problem of obtaining exact goodness-of-fit tests for certain parametric NHPPs, using a method based on Monte Carlo simulation conditional on sufficient statistics. A closely related way of obtaining exact confidence intervals in parametri...
متن کاملparty: A Laboratory for Recursive Partytioning
The party package (Hothorn, Hornik, and Zeileis 2006) aims at providing a recursive part(y)itioning laboratory assembling various highand low-level tools for building tree-based regression and classification models. This includes conditional inference trees (ctree), conditional inference forests (cforest) and parametric model trees (mob). At the core of the package is ctree, an implementation o...
متن کاملContext models on sequences of covers
We present a class of models that, via a simple construction, enables exact, incremental, non-parametric, polynomial-time, Bayesian inference of conditional measures. The approach relies upon creating a sequence of covers on the conditioning variable and maintaining a different model for each set within a cover. Inference remains tractable by specifying the probabilistic model in terms of a ran...
متن کاملDecomposing Local Probability Distributions in Bayesian Networks for Improved Inference and Parameter Learning
Amajor difficulty in building Bayesian network models is the size of conditional probability tables, which grow exponentially in the number of parents. One way of dealing with this problem is through parametric conditional probability distributions that usually require only a linear number of parameters in the number of parents. In this paper we introduce a new class of parametric models, the p...
متن کاملA Combined Approach to the Inference of Conditional Factor Models
This paper develops a new methodology for estimating and testing conditional factor models in finance. We propose a two-stage procedure that naturally unifies the two existing approaches in the finance literature–the parametric approach and the nonparametric approach. Our combined approach possesses important advantages over both methods. Using our two-stage combined estimator, we derive new te...
متن کامل