Home > MRFLearning > PMM.m

PMM Learning for boltzmann m/c The Pseudo Moment Matching method to estimate parameters of a boltzmann m/c Inputs: data: (N X V) matrix containing the data samples. N is the total number of samples and V is the total number of nodes Each node can take values from {+1,-1} or {0,1}. A: (V X V) adjacency matrix defining the graph structure Returns: model: (1X1) struct array with fields N: the number of nodes A: adjacency matrix b: the biases ( PL estimates ) w: the edge weights ( PL estimates ) The node value representation ( +1/-1 or 0/1) intended by the user is detected from the training samples. If using +1/-1, the data is first mapped to 0/1 and the model params are learned. These learned params are then mapped back to the +1/-1 case. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Generated on Wed 28-Sep-2005 04:10:02 by