Table 1.

List of notation.

D	Input space dimensionality
N	Number of training instances
R	Projected subspace dimensionality
L	Number of output labels
X ∈ ℝD×N	Data matrix (i.e., collection of training instances)
Q ∈ ℝD×R	Projection matrix (i.e., dimensionality reduction parameters)
Φ ∈ ℝD×R	Priors for projection matrix (i.e., precision priors)
Z ∈ ℝR×N	Projected data matrix (i.e., low-dimensional representation of instances)
W ∈ ℝL×R	Weight matrix (i.e., classification parameters for labels)
Ψ ∈ ℝL×R	Priors for weight matrix (i.e., precision priors)
b ∈ ℝL	Bias vector (i.e., bias parameters for labels)
λ ∈ ℝL	Priors for bias vector (i.e., precision priors)
T ∈ ℝL×N	Auxiliary matrix (i.e., discriminant outputs)
Y ∈ {±1}L×N	Label matrix (i.e., true labels of training instances)