Multi-Dictionary Tensor Decomposition

Tensor decomposition methods are popular tools for analysis of multi-way datasets from the social media, healthcare, spatio-temporal domains, and others. Widely adopted models such as Tucker and canonical polyadic decomposition (CPD) follow a data-driven philosophy: they decompose a tensor into factors that approximate the observed data well. In some cases side information is available about the tensor modes. For example, in a temporal user-item purchases tensor a user influence graph, an item similarity graph, and knowledge about seasonality or trends in the temporal mode may be available. Such side information may enable more succinct and interpretable tensor decomposition models and improved quality in downstream tasks. We propose a framework for Multi-Dictionary Tensor Decomposition (MDTD) which takes advantage of prior structural information about tensor modes in the form of coding dictionaries to obtain sparsely coded tensor factors. We derive a general optimization algorithm for MDTD that handles both complete inputs and inputs with missing values. MDTD handles large sparse tensors typical in many real-world application domains. We experimentally demonstrate its utility in both synthetic and real-world datasets. It learns more concise models than dictionary-free counterparts and improves (i) reconstruction quality (up to 60% smaller models coupled with reduced representation error); (ii) missing values imputation quality (two-fold MSE reduction with up to orders of magnitude time savings) and (iii) the estimation of the tensor rank. MDTD’s quality improvements do not come with a running time premium: it can decompose 19GB datasets in less than a minute. It can also impute missing values in sparse billion-entry tensors more accurately and scalably than state-of-the-art competitors.