Probabilistic non-negative matrix factorisation and extensions
نویسنده
چکیده
Matrix factorisation models have had an explosive growth in popularity in the last decade. It has become popular due to its usefulness in clustering and missing values prediction. We review the main literature for matrix factorisation, focusing on nonnegative matrix factorisation and probabilistic approaches. We also consider several extensions: matrix tri-factorisation, Tensor factorisation, Tucker decomposition, and data fusion.
منابع مشابه
Efficient Bayesian Community Detection using Non-negative Matrix Factorisation
Identifying overlapping communities in networks is a challenging task. In this work we present a novel approach to community detection that utilises the Bayesian non-negative matrix factorisation (NMF) model to produce a probabilistic output for node memberships. The scheme has the advantage of computational efficiency, soft community membership and an intuitive foundation. We present the perfo...
متن کاملAligning words using matrix factorisation
Aligning words from sentences which are mutual translations is an important problem in different settings, such as bilingual terminology extraction, Machine Translation, or projection of linguistic features. Here, we view word alignment as matrix factorisation. In order to produce proper alignments, we show that factors must satisfy a number of constraints such as orthogonality. We then propose...
متن کاملAlgorithme EM en ligne simulé pour la factorisation non-négative probabiliste
This paper describes a general purpose algorithm for online estimation in probabilistic latent factor models. The algorithm called simulated online EM is an extension of the online EM algorithm of Cappé & Moulines (2009) based on the use of MCMC simulations to approximate individual E-steps. The algorithm is applied to perform unsupervised analysis of audio signals using a probabilistic version...
متن کاملFast Bayesian Non-Negative Matrix Factorisation and Tri-Factorisation
Nonnegative matrix factorisation and tri-factorisation Nonnegative matrix factorisation (NMF) and tri-factorisation (NMTF) methods decompose a given matrix R into two or three smaller matrices so that R ≈ UV T or R ≈ FSG , respectively. Schmidt, Winther and Hansen (2009) introduced a Bayesian version of nonnegative matrix factorisation (left), which we extend to matrix tri-factorisation (right)...
متن کاملNon-negative tensor factorisation of modulation spectrograms for monaural sound source separation
This paper proposes an algorithm for separating monaural audio signals by non-negative tensor factorisation of modulation spectrograms. The modulation spectrogram is able to represent redundant patterns across frequency with similar features, and the tensor factorisation is able to isolate these patterns in an unsupervised way. The method overcomes the limitation of conventional non-negative ma...
متن کامل