Changes between Version 14 and Version 15 of MatrixFactorization
 Timestamp:
 07/20/11 16:44:22 (3 years ago)
Legend:
 Unmodified
 Added
 Removed
 Modified

MatrixFactorization
v14 v15 17 17 Reference: (Brunet, 2004). 18 18 19 === snmf ===19 === snmf (snmf/r and snmf/l) === 20 20 Sparse NMF based on alternating nonnegativity constrained least squares, solved by a fast nonnegativity constrained least squares. Sparseness imposed on the left, right factor. It is meant to be very fast compared to other approaches. [[BR]] 21 21 Reference: (Kim, 2007). … … 34 34 35 35 === pmf === 36 Probabilistic MF. Model whichscales linearly with the number of observations and performs well on large, sparse, imbalanced datasets.[[BR]]36 Probabilistic MF. PMF model scales linearly with the number of observations and performs well on large, sparse, imbalanced datasets.[[BR]] 37 37 Reference: (Salakhutdinov, 2007). 38 39 === Additional: bpmf === 40 Bayesian PMF model that implements Gibbs sampler. In Bayesian PMF model capacity is controlled automatically by integrating over all model parameters and hyperparameters. Bayesian PMF can be efficiently trained using MCMC methods. This model achieves significantly higher prediction accuracy than PMF models trained using MAP estimation. This algorithm has been used by winning and top ranked teams in KDD Cup 2011. [[BR]] 41 Reference: (Salakhutdinov, Mnih, 2008). 38 42 39 43 === psmf === … … 53 57 * fixed 54 58 * nndsvd 59 * pmf 55 60 * algorithm specific 56 61 … … 139 144 * Ochs, M. F., Kossenkov A. V. NIH Public Access. Methods, Methods Enzymol., 2009, 5977. 140 145 * Z. Zhang, T. Li, C. H. Q. Ding, X. Zhang: Binary Matrix Factorization with Applications. ICDM 2007. 146 * Salakhutdinov, R., Mnih, A. Bayesian probabilistic matrix factorization using Markov chain Monte Carlo. ICML 2008, 880887.