Ranking evaluation of institutions based on a Bayesian network having a latent variable

作者：

Highlights：

• Institutions will be ranked based on a set of performance indicators.

• A Bayesian network is constructed having a latent variable.

• Linear Gaussian models are used to express the causal relation among variables.

• An interval estimate of the ranking of an institution is obtained.

• A real case and a simulation study are considered.

摘要

•Institutions will be ranked based on a set of performance indicators.•A Bayesian network is constructed having a latent variable.•Linear Gaussian models are used to express the causal relation among variables.•An interval estimate of the ranking of an institution is obtained.•A real case and a simulation study are considered.

论文关键词：Ranking estimation,Linear Gaussian model,Structure learning,Gibbs sampling,Multiple search,Causal discovery

论文评审过程：Received 5 April 2012, Revised 29 April 2013, Accepted 22 May 2013, Available online 31 May 2013.

论文官网地址：https://doi.org/10.1016/j.knosys.2013.05.010