For each RBM layer, an RBM class is initialized. This class extends the UnsupervisedModel and Model classes:
Details of the RBM class __init__(..) function are specified in the following code:
class RBM(UnsupervisedModel): """Restricted Boltzmann Machine implementation using TensorFlow. The interface of the class is sklearn-like. """ def __init__( self, num_hidden, visible_unit_type='bin', name='rbm', loss_func='mse', learning_rate=0.01, regcoef=5e-4, regtype='none', gibbs_sampling_steps=1, batch_size=10, num_epochs=10, stddev=0.1): """Constructor. :param num_hidden: number of hidden units :param loss_function: type of loss function ...
