TY - JOUR
T1 - Mean–field variational approximate Bayesian inference for latent variables models
AU - Consonni, Guido
PY - 2007
Y1 - 2007
N2 - The ill-posed nature of missing variable models offers a challenging testing ground for new computational techniques. This is the
case for the mean-field variational Bayesian inference. The behavior of this approach in the setting of the Bayesian probit model
is illustrated. It is shown that the mean-field variational method always underestimates the posterior variance and, that, for small
sample sizes, the mean-field variational approximation to the posterior location could be poor.
AB - The ill-posed nature of missing variable models offers a challenging testing ground for new computational techniques. This is the
case for the mean-field variational Bayesian inference. The behavior of this approach in the setting of the Bayesian probit model
is illustrated. It is shown that the mean-field variational method always underestimates the posterior variance and, that, for small
sample sizes, the mean-field variational approximation to the posterior location could be poor.
KW - Latent variable model
KW - Mean-field variational method
KW - Latent variable model
KW - Mean-field variational method
UR - http://hdl.handle.net/10807/10632
UR - http://dx.medra.org/10.1016/j.csda.2006.10.028
U2 - 10.1016/j.csda.2006.10.028
DO - 10.1016/j.csda.2006.10.028
M3 - Article
SN - 0167-9473
VL - 52
SP - 790
EP - 798
JO - COMPUTATIONAL STATISTICS & DATA ANALYSIS
JF - COMPUTATIONAL STATISTICS & DATA ANALYSIS
ER -