Rank regularization and Bayesian inference for tensor completion and extrapolation

Abstract

A novel regularizer of the PARAFAC decomposition factors capturing the tensor's rank is proposed in this paper, as the key enabler for completion of three-way data arrays with missing entries. Set in a Bayesian framework, the tensor completion method incorporates prior information to enhance its smoothing and prediction capabilities. This probabilistic approach can naturally accommodate general models for the data distribution, lending itself to various fitting criteria that yield optimum estimates in the maximum-a-posteriori sense. In particular, two algorithms are devised for Gaussian- and Poisson-distributed data, that minimize the rank-regularized least-squares error and Kullback-Leibler divergence, respectively. The proposed technique is able to recover the "ground-truth'' tensor rank when tested on synthetic data, and to complete brain imaging and yeast gene expression datasets with 50% and 15% of missing entries respectively, resulting in recovery errors at -10dB and -15dB.

12 pages, submitted to IEEE Transactions on Signal Processing

References in corpus (3)

• On Tensors, Sparsity, and Nonnegative Factorizations
• A New Approach to Collaborative Filtering: Operator Estimation with Spectral Regularization
• In-network Sparsity-regularized Rank Minimization: Algorithms and Applications

Cited by in corpus (23)

• Integrated Sensing and Communication with Massive MIMO: A Unified Tensor Approach for Channel and Target Parameter Estimation
• Subspace Learning and Imputation for Streaming Big Data Matrices and Tensors
• Network Volume Anomaly Detection and Identification in Large-scale Networks based on Online Time-structured Traffic Tensor Tracking
• Phased Array-Based Sub-Nyquist Sampling for Joint Wideband Spectrum Sensing and Direction-of-Arrival Estimation
• A Unified Algorithmic Framework for Block-Structured Optimization Involving Big Data
• An Iterative Reweighted Method for Tucker Decomposition of Incomplete Multiway Tensors
• Identification of Overlapping Communities via Constrained Egonet Tensor Decomposition
• Bayesian Tensorized Neural Networks with Automatic Rank Selection
• Coupled Graphs and Tensor Factorization for Recommender Systems and Community Detection
• Generalization error bounds for kernel matrix completion and extrapolation
• Tensor Robust Principal Component Analysis: Better recovery with atomic norm regularization
• Tracking Tensor Subspaces with Informative Random Sampling for Real-Time MR Imaging
• Alternating minimization algorithms for graph regularized tensor completion
• Bayesian Methods in Tensor Analysis
• A Unified Framework for Coupled Tensor Completion
• Adaptive Anomaly Detection in Network Flows with Low-Rank Tensor Decompositions and Deep Unrolling
• Channel Estimation for Millimeter Wave Multiuser MIMO Systems via PARAFAC Decomposition
• Tuning Free Rank-Sparse Bayesian Matrix and Tensor Completion with Global-Local Priors
• Low-Rank Tensor Decomposition-Aided Channel Estimation for Millimeter Wave MIMO-OFDM Systems
• Variational Bayesian inference for CP tensor completion with side information
• Estimating Traffic and Anomaly Maps via Network Tomography
• Spectral Compressed Sensing via CANDECOMP/PARAFAC Decomposition of Incomplete Tensors
• Learning the Sparse and Low Rank PARAFAC Decomposition via the Elastic Net