Graphical Models

From RedwoodCenter
Jump to navigationJump to search
Models Marginal distribution over 'source' vectors
(before an observation)
Conditional distributions over 'source' vectors
(after observation)
Square Independent
(by assumption)
Independent
(because deterministic)
Causal
overcomplete
Independent
(by assumption)
Dependent
(explaining away)
Energy-based
overcomplete
Dependent
(rejecting-away)
Independent
(because deterministic)

from Energy-Based Models for Sparse Overcomplete Representations Teh, Y. W, Welling, M., Osindero, S. and Hinton G. E., Journal of Machine Learning Research, 4, 2003 pp 1235-1260 [[1]]