Learn R Programming

icmm (version 1.2)

Empirical Bayes Variable Selection via ICM/M Algorithm

Description

Empirical Bayes variable selection via ICM/M algorithm for normal, binary logistic, and Cox's regression. The basic problem is to fit high-dimensional regression which sparse coefficients. This package allows incorporating the Ising prior to capture structure of predictors in the modeling process. More information can be found in the papers listed in the URL below.

Copy Link

Version

Install

install.packages('icmm')

Monthly Downloads

182

Version

1.2

License

GPL (>= 2)

Homepage

https://www.researchgate.net/publication/279279744_Selecting_massive_variables_using_an_iterated_conditional_modesmedians_algorithm

Maintainer

Vitara Pungpapong

Last Published

May 26th, 2021

Functions in icmm (1.2)

Empirical Bayes Variable Selection

initbetaGaussian

Initial values for the regression coefficients used in example for running ICM/M algorithm in normal linear regression model

Local posterior probability estimation

Initial values for the regression coefficients used in example for running ICM/M algorithm in Cox's model

Mixing weight estimation.

Linear structure of predictors

Simulated data from the binary logistic regression model

initbetaBinomial

Initial values for the regression coefficients used in example for running ICM/M algorithm in binary logistic model

Empirical Bayes Variable Selection via ICM/M

Hyperparameter estimation for alpha.

Local posterior probability estimation.

get.pseudodata.binomial

Obtain pseudodata based on the binary logistic regression model.

Obtain model coefficient without assuming prior on structure of predictors.

Estimate posterior probability of mixing weight.

Standard deviation estimation.

Hyperparameter estimation for a and b.

Simulated data from the normal linear regression model

get.pseudodata.cox

Obtain pseudodata based on the Cox's regression model.

Obtain a regression coefficient when assuming Ising prior (with structured predictors).

Simulated data from Cox's regression model