2. 您的位置
3. 首页
4. 会议论文
5. Leveraging Algorithms for Super-Large Sample Logistic Regression

Leveraging Algorithms for Super-Large Sample Logistic Regression

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

【摘 要】

：

For massive data with super-large sample size,it is computationally infeasible to obtain maximum likelihood estimates for unknown parameters,especially when the estimators do not have closed-form solu

【作 者】

：

Ping Ma 

【机 构】

：

University of Georgia,Athens,GA 30602,USA

【出 处】

：

上海交通大学

【发表日期】

：

2016年4期

【关键词】

：

subsampling asymptotics maximum likelihood estimate large sample theory mean squ 

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

　　For massive data with super-large sample size,it is computationally infeasible to obtain maximum likelihood estimates for unknown parameters,especially when the estimators do not have closed-form solutions.

其他文献

On dimensionality effects in linear discriminant analysis for large dimensional data

We study the asymptotic results of linear discriminant analysis(LDA)in large dimensional data where the observation dimension p,is of the same order of magnitude as the sample size n.

会议

Dimensionality EffectLinear Discriminant AnalysisMisclassification Error Rate

Reference-free Learning with Multiple Metagenomic Samples

The major goals of metagenomics are to identify and study the entire collection of microbial species in a set of targeted samples through sequencing bulk DNA extracted directly from the samples.

会议

MetagenomicsBinningMixture ModelMultinomialUnsupervised Learning

Bayesian Nonparametric Monotone Regression of Dynamic Latent Traits in Item Response Models

The parametric methods,such as autoregressive models or latent growth modeling are usually inflexible to model the dependence and nonlinear effects among the changes of latent traits whenever the time

会议

Bayesian NonparametricMonotonic RegressionDynamic Item Response Model

SPOT:Sparse Optimal Transformations for High Dimensional Variable Selection and Exploratory Regressi

We develop a novel method called SParse Optimal Transformations(SPOT)to simultaneously select important variables and explore relationships between the response and predictor variables in high dimensi

会议

Optimal transformationnonparametric regressionvariable selection

Bayesian influence analysis of nonlinear reproductive dispersion mixed models for longitudinal data

This paper develops three Bayesian case influence measures including the -divergence,Cooks posterior mode distance and Cooks posterior mean distance to identify influential observations,and a Bayesia

会议

Bayesian case influenceBayesian local influencemissing covariatesnonlinear re

Semiparametric Inference for the Proportional Mean Residual Life Model with Right-Censored Length-Bi

We propose a semiparametric inference approach for the proportional mean residual life model with right-censored length-biased data,which arise frequently in observational studies,espe-cially in epide

会议

Proportional mean residual modelLength-biased dataDependent censoringEstimat-

Coping with missing data in Phase Ⅲ pivotal registration trials:Tolvaptan in Subjects with kidney di

Missing data cause challenging issues,particularly in Phase Ⅲ registration trials,as highlighted by the European Medicines Agency(EMA)and the US National Research Council(NRC)..We explore,as a case st

会议

missing datamissing not at randommultiple imputationstipping point analysis

Nonparametric estimator of strictly monotone regression function for longitudinal data

Recently,longitudinal data have been widely recorded over sparse or dense time points in various imaging studies.

会议

Nonparametric estimationKernel smoothingIsotone regressionSparse and dense lo

One dimensional free fermions at finite temperature and the MNS matrix model

In this talk we discuss the local statistics of free fermions on the real line,in a quadratic potential well,and at finite temperature.This free fermion model is equivalent to a matrix model defined b

会议

free fermionsMNS matrix modelfinite temperature

A SIEVE SEMIPARAMETRIC MAXIMUM LIKELIHOOD ESTIMATION METHOD FOR ACCELERATED HAZARD MODEL WITH INTERV

The accelerated hazards(AH)model has been developed for right censored survival data to apply in situations where the proportional hazards model and other commonly used semiparametric models are not s

会议

interval censoringaccelerated hazards modeltwo-step optimization