Kernel Probabilistic Dependent-Independent Canonical Correlation Analysis

Algorithm 1

PDICCA algorithm.

(1)	Assume that and are fixed and marginalized over and to get
	and
	(i) Update the parameters using
	Here , , and is a block-diagonal matrix that consist of and . The is the joint sample covariance matrix.
(2)	Marginalize over to get
	(i) Update with
	where and . And is the sample covariance of .
	(ii) Update using
	where is the dimensionality of , and is the new value just updated.
	Repeat the above two substeps for parameters related to y, replacing all subscripts x with y.