Gibbs ensembles for nearly compatible and incompatible conditional models

TY - JOUR

T1 - Gibbs ensembles for nearly compatible and incompatible conditional models

AU - Chen, Shyh Huei

AU - Ip, Edward H.

AU - Wang, Yuchung J.

PY - 2011/4/1

Y1 - 2011/4/1

N2 - The Gibbs sampler has been used exclusively for compatible conditionals that converge to a unique invariant joint distribution. However, conditional models are not always compatible. In this paper, a Gibbs sampling-based approachusing the Gibbs ensembleis proposed for searching for a joint distribution that deviates least from a prescribed set of conditional distributions. The algorithm can be easily scalable, such that it can handle large data sets of high dimensionality. Using simulated data, we show that the proposed approach provides joint distributions that are less discrepant from the incompatible conditionals than those obtained by other methods discussed in the literature. The ensemble approach is also applied to a data set relating to geno-polymorphism and response to chemotherapy for patients with metastatic colorectal cancer.

AB - The Gibbs sampler has been used exclusively for compatible conditionals that converge to a unique invariant joint distribution. However, conditional models are not always compatible. In this paper, a Gibbs sampling-based approachusing the Gibbs ensembleis proposed for searching for a joint distribution that deviates least from a prescribed set of conditional distributions. The algorithm can be easily scalable, such that it can handle large data sets of high dimensionality. Using simulated data, we show that the proposed approach provides joint distributions that are less discrepant from the incompatible conditionals than those obtained by other methods discussed in the literature. The ensemble approach is also applied to a data set relating to geno-polymorphism and response to chemotherapy for patients with metastatic colorectal cancer.

KW - Conditionally specified distribution

KW - Ensemble method

KW - Gibbs sampler

KW - Linear programming

KW - Odds ratio

UR - http://www.scopus.com/inward/record.url?scp=78650823497&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78650823497&partnerID=8YFLogxK

U2 - https://doi.org/10.1016/j.csda.2010.11.006

DO - https://doi.org/10.1016/j.csda.2010.11.006

M3 - Article

SN - 0167-9473

VL - 55

SP - 1760

EP - 1769

JO - Computational Statistics and Data Analysis

JF - Computational Statistics and Data Analysis

IS - 4

ER -