ContrastiveDivergence

Learn a matrix of data. Each row represents a single training set. For example t = [[1,0],[0,1]] will train the rbm to recognize [1,0] and [0,1] If a Matrix is too large I recommend splitting it into smaller matrices, but if they are reasonably small, this is a fast way to "simultaneously" train multiple inputs and should generable be used unless the matrices are just too large