This work presents simple and fast structured Bayesian learning for matrix and tensor factorization models. An unblocked Gibbs sampler is proposed for factorization machines (FM) which are a general class of latent variable models subsuming matrix, tensor and many other factorization models. We empirically show on the large Netflix challenge dataset that Bayesian FM are fast, scalable and more ...

Matrix factorization or factor analysis is an important task helpful in the analysis of high dimensional real world data. There are several well known methods and algorithms for factorization of real data but many application areas including information retrieval, pattern recognition and data mining require processing of binary rather than real data. Unfortunately, the methods used for real mat...

Factorsynth is a new software tool, developed in the Max environment, that implements sound processing based on matrix factorization techniques. In particular, Nonnegative Matrix Factorization is applied to the input sounds, which produces a set of temporal and spectral components that can be then freely manipulated and combined to produce new sounds. Based on a simple graphical interface that ...

In the literature, one often finds calculations of (e, ep) reactions at GeV energies using the factorization approach. Factorization implies that the differential cross section can be written as the product of an off-shell electron-proton cross section and a distorted missing momentum distribution. While this factorization appears in the non-relativistic plane wave impulse approximation, it is ...

Transverse-momentum dependent (TMD) light-cone wave functions of a light meson are important ingredients in the TMD QCD factorization of exclusive processes. This factorization allows one conveniently resum Sudakov logarithms appearing in collinear factorization. The TMD light-cone wave functions are not simply related to the standard light-cone wave functions in collinear factorization by inte...

In the literature, one often finds calculations of (e, ep) reactions at GeV energies using the factorization approach. Factorization implies that the differential cross section can be written as the product of an off-shell electron-proton cross section and a distorted missing momentum distribution. While this factorization appears in the non-relativistic plane wave impulse approximation, it is ...

It is today accepted that matrix factorization models allow a high quality of rating prediction in recommender systems. However, a major drawback of matrix factorization is its static nature that results in a progressive declining of the accuracy of the predictions after each factorization. This is due to the fact that the new obtained ratings are not taken into account until a new factorizatio...

We examine the problem of producing the optimal evaluation order for queries containing joins, selections, and maps. Specifically, we look at the case where common subexpressions involving expensive UDF calls can be factored out. First, we show that ignoring factorization during optimization can lead to plans that are far off the best possible plan: the difference in cost between the best plan ...

The ILU factorization is one of the most popular preconditioners for the Krylov subspace method, alongside the GMRES. Properties of the preconditioner derived from the ILU factorization are relayed onto the dropping rules. Recently, Zhang et al. [Numer. Linear. Algebra. Appl., Vol. 19, pp. 555–569, 2011] proposed a Flexible incomplete Cholesky (IC) factorization for symmetric linear systems. Th...

It is today accepted that matrix factorization models allow a high quality of rating prediction in recommender systems. However, a major drawback of matrix factorization is its static nature that results in a progressive declining of the accuracy of the predictions after each factorization. This is due to the fact that the new obtained ratings are not taken into account until a new factorizatio...