Files
test/machine_learning/course2/assignment4/hidden_state_to_visible_probabilities.m

10 lines
674 B
Matlab

function visible_probability = hidden_state_to_visible_probabilities(rbm_w, hidden_state)
% <rbm_w> is a matrix of size <number of hidden units> by <number of visible units>
% <hidden_state> is a binary matrix of size <number of hidden units> by <number of configurations that we're handling in parallel>.
% The returned value is a matrix of size <number of visible units> by <number of configurations that we're handling in parallel>.
% This takes in the (binary) states of the hidden units, and returns the activation probabilities of the visible units, conditional on those states.
deltaE = rbm_w' * hidden_state;
visible_probability = logistic(deltaE);
end