A new robust regression model:Type Ⅱ multivariate t distribution with applications

来源 :上海交通大学 | 被引量 : 0次 | 上传用户：hzq1989aa

【摘要】

：

【作者】

：

Chi ZHANG Guo-Liang TIAN

【机构】

：

the University of Hong Kong,999077,Hong Kong Sout

【出处】

：

上海交通大学

【发表日期】

：

2016年4期

【关键词】

：

Expectation/conditional maximization algorithm multivariate t distribution multi

下载到本地 , 更方便阅读

下载此文赞助VIP

声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架

论文部分内容阅读

　　Motivated by a real data analysis,we in this paper propose a new multivariate t(MVT)distribution by allowing variable degrees of freedom for each univariate component.This new distribution is called Type Ⅱ MVT distribution,which possesses several remarkable features including(1)all components follow univariate t-distributions with possible different degrees of freedom,(2)it could include components following the multivariate normal distributions when the corresponding degrees of freedom approach to infinity,and(3)it contains the product of independent t-distributions as a particular case.

其他文献

Novel diagnostic accuracy analysis for competing risks outcomes with ROC surface

Many medical conditions are marked by a sequence of events or statuses that are associated with continuous changes in some biomarkers.However,few works have been done to assess the overall accuracy of

会议

ClassificationConcordanceInverse probability weightingU statisticVolume unde

Modeling Non-Gaussian Time Series with Nonparametric Bayesian Model

We present a class of Bayesian copula models whose major components are the marginal(limiting)distribution of a stationary time series and the internal dynamics of the series.We argue that these are t

会议

Autoregressive processCopula modelGARCHprobability integral transformation

ESTIMATING OPTIMAL DYNAMIC TREATMENT REGIMES WITH SHARED DECISION RULES

A dynamic treatment regime consists of decision rules that recommend how to individualize treatment to patients based on available treatment and covariate history.In many scientific domains,these deci

会议

dynamic treatment regimesQ-learningshared parametersSTAR*D

Regression-Based Covariance Functions for Nonstationary Spatial Modeling

In Gaussian process(GP)models for spatial data,the covariance function is typically assumed to belong to a parametric class of stationary covariance models.Various authors have proposed methods for re

会议

Bayesian modelingGaussian processesNonstationarityPrecipitationSpatial stati

Pseudo-time reconstruction and evaluation in single-cell RNA-seq analysis

When analyzing single-cell RNA-seq data,constructing a pseudo-temporal path to order cells based on the gradual transition of their transcriptomes is a useful way to study gene expression dynamics in

会议

High-throughput sequencingsingle cellgenomicsminimum spanning treeRNA-seq

Method of moments in analysis of algorithms:a selected survey

The method of moments is among the very first approaches used to establish the classical central limit theorem and has a long history in the literature.In modern times its use has been largely oversha

会议

Method of momentsdiscrete probabilityanalysis of algorithmsrandom treeslimit

Estimating probability of presence from presence-only data

Accurately predicting species distribution is essential for many ecological and conservation problems.For many species,the most commonly available type of data is the presence-only(presence-background

会议

absolute probability of presenceLK methodLI methodspecies distribution modell

Composite likelihood inference on stochastic block model for big networks

Hi-C experiment allows the measurements of the frequencies of physical contacts among pairs of genomic loci at a genome-wide scale.A fundamental step of analyzing such data is to detect and model the

会议

Stochastic Block ModelComposite likelihoodHi-C data

Statistical Implications of Extrapolating the Overall Result to the Target Region in Multi-Regional

The one of principles described in ICH E9 is that only results obtained from pre-specified statistical methods in protocol is regarded as confirmatory evidences.However,in multi-regional clinical tria

会议

consistencytype Ⅰ error ratepowerintrinsic factorextrinsic factor

Approaches for Massive Spatial Data and Applications in Remote Sensing

With the development of new remote sensing technology,large or even massive spatial datasets from Earth observation become available.Statistical analysis of such data is challenging.We proposes a semi

会议

Basis functionGaussian Markov random fieldRemote sensingSemiparametric method

A new robust regression model:Type Ⅱ multivariate t distribution with applications

其他学术论文