A Nonconvex Low-Rank Tensor Completion Model for Spatiotemporal Traffic Data Imputation

Abstract

Sparsity and missing data problems are very common in spatiotemporal traffic data collected from various sensing systems. Making accurate imputation is critical to many applications in intelligent transportation systems. In this paper, we formulate the missing data imputation problem in spatiotemporal traffic data in a low-rank tensor completion (LRTC) framework and define a novel truncated nuclear norm (TNN) on traffic tensors of location$\times$day$\times$time of day. In particular, we introduce an universal rate parameter to control the degree of truncation on all tensor modes in the proposed LRTC-TNN model, and this allows us to better characterize the hidden patterns in spatiotemporal traffic data. Based on the framework of the Alternating Direction Method of Multipliers (ADMM), we present an efficient algorithm to obtain the optimal solution for each variable. We conduct numerical experiments on four spatiotemporal traffic data sets, and our results show that the proposed LRTC-TNN model outperforms many state-of-the-art imputation models with missing rates/patterns. Moreover, the proposed model also outperforms other baseline models in extreme missing scenarios.

References in corpus (5)

• Linearized Alternating Direction Method with Adaptive Penalty for Low-Rank Representation
• Smooth PARAFAC Decomposition for Tensor Completion
• Generalized Nonconvex Nonsmooth Low-Rank Minimization
• Bayesian Temporal Factorization for Multidimensional Time Series Prediction
• Factor Group-Sparse Regularization for Efficient Low-Rank Matrix Recovery

Cited by in corpus (12)

• Low-Rank Autoregressive Tensor Completion for Spatiotemporal Traffic Data Imputation
• ImputeFormer: Low Rankness-Induced Transformers for Generalizable Spatiotemporal Imputation
• Scalable Low-Rank Tensor Learning for Spatiotemporal Traffic Data Imputation
• Truncated tensor Schatten p-norm based approach for spatiotemporal traffic data imputation with complicated missing patterns
• Physics-Informed Deep Learning For Traffic State Estimation: A Survey and the Outlook
• Laplacian Convolutional Representation for Traffic Time Series Imputation
• Low-Rank Autoregressive Tensor Completion for Multivariate Time Series Forecasting
• A generalizable framework for low-rank tensor completion with numerical priors
• Spatiotemporal Implicit Neural Representation as a Generalized Traffic Data Learner
• Low-Rank Hankel Tensor Completion for Traffic Speed Estimation
• Incremental Bayesian tensor learning for structural monitoring data imputation and response forecasting
• A Biased Deep Tensor Factorization Network For Tensor Completion