High performance Gibbs sampling for IRT models using row-wise decomposition

DOI10.5402/2012/264040MaRDI QIDQ355520zbMATH OpenOpenAlexWikidataFDO

Authors Yanyan Sheng, Mona Rahimi

Publication date 25 July 2013

Published in ISRN Computational Mathematics (Search for Journal in Brave)

Full work available at URL https://doi.org/10.5402/2012/264040

Mathematics Subject Classification ID

Applications of statistics to psychology (62P15) Memory and learning in psychology (91E40) Communication theory (94A05) Sampling theory in information and communication theory (94A20)

Review text

Summary: Item response theory (IRT) is a popular approach used for addressing statistical problems in psychometrics as well as in other fields. The fully Bayesian approach for estimating IRT models is computationally expensive. This limits the use of the procedure in real applications. In an effort to reduce the execution time, a previous study shows that high performance computing provides a solution by achieving a considerable speedup via the use of multiple processors. Given the high data dependencies in a single Markov chain for IRT models, it is not possible to avoid communication overhead among processors. This study is to reduce communication overhead via the use of a row-wise decomposition scheme. The results suggest that the proposed approach increased the speedup and the efficiency for each implementation while minimizing the cost and the total overhead. This further sheds light on developing high performance Gibbs samplers for more complicated IRT models.

Recommendations

• A GPU-based Gibbs sampler for a unidimensional IRT model
• Gibbs sampling method in multidimensional two parameter logistic item response model
• Exploring the posterior of a hierarchical IRT model for item effects
• Gibbs samplers for logistic item response models via the Pólya-gamma distribution: a computationally efficient data-augmentation strategy
• A Gibbs sampler for the multidimensional item response model

Cites work

• Bayesian estimation of item response curves
• scientific article; zbMATH DE number 472921 (Why is no real title available?)
• scientific article; zbMATH DE number 806780 (Why is no real title available?)
• scientific article; zbMATH DE number 814821 (Why is no real title available?)
• scientific article; zbMATH DE number 3297790 (Why is no real title available?)
• Item Response Theory
• Markov chains for exploring posterior distributions. (With discussion)
• Statistical theory for logistic mental test models with a prior distribution of ability
• Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images

Cited in

(4)

• A GPU-based Gibbs sampler for a unidimensional IRT model
• Real-time Bayesian parameter estimation for item response models
• Using Stan to implement Bayesian parameter estimation of IRT models
• py-irt: A Scalable Item Response Theory Library for Python

Describes a project that uses

Uses Software

• GSL

This page was built for publication: High performance Gibbs sampling for IRT models using row-wise decomposition

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q355520)