TY - GEN
T1 - Bayesian supervised learning with non-Gaussian latent variables
AU - Lyu, Siwei
PY - 2013
Y1 - 2013
N2 - We describe a Bayesian learning scheme for the hierarchal Bayesian linear model, which is based on the Gaussian scale mixture (GSM) modeling of the distribution of the latent variable. The proposed method takes advantage of the hierarchal Gaussian structure for a simple Monte-Carlo sampling algorithm. Particularly, with a single hidden scale parameter controlling the distribution of the latent variables, it leads to an efficient algorithm without explicit matrix inversion.
AB - We describe a Bayesian learning scheme for the hierarchal Bayesian linear model, which is based on the Gaussian scale mixture (GSM) modeling of the distribution of the latent variable. The proposed method takes advantage of the hierarchal Gaussian structure for a simple Monte-Carlo sampling algorithm. Particularly, with a single hidden scale parameter controlling the distribution of the latent variables, it leads to an efficient algorithm without explicit matrix inversion.
KW - Bayesian learning
KW - Gaussian scale mixtures
KW - latent variable models
UR - https://www.scopus.com/pages/publications/84889602505
U2 - 10.1109/ChinaSIP.2013.6625424
DO - 10.1109/ChinaSIP.2013.6625424
M3 - Conference contribution
AN - SCOPUS:84889602505
SN - 9781479910434
T3 - 2013 IEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2013 - Proceedings
SP - 659
EP - 663
BT - 2013 IEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2013 - Proceedings
T2 - 2013 IEEE China Summit and International Conference on Signal and Information Processing, ChinaSIP 2013
Y2 - 6 July 2013 through 10 July 2013
ER -