
Saunder August 4, 2015 16:30 K23081˙C002
Visual Features Extraction
23
activation [103]. The output weights W ∈
R
K ×N
and biases b ∈
R
K
so that
the feature mapping f is
f (x) = g(Wx + b) (2.17)
where g (z) = 1/(1 +exp(−z)) is the logistic sigmoid function.
Sparse restricted Boltzmann machine. A restricted Boltzmann machine (RBM)
is a generative model. The RBM is likely the most popular subclass of
Boltzmann machine [171]. The RBM can be said to form a bipartite graph
with the visibles and the hidden forming two layers of vertices in the graph
(and no connection between units of the same layer) with K binary hidden
variables. Sparse RBMs can be trained using ...