Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Missing Slice Recovery for Tensors Using a Low-rank Model in Embedded Space

About

Let us consider a case where all of the elements in some continuous slices are missing in tensor data. In this case, the nuclear-norm and total variation regularization methods usually fail to recover the missing elements. The key problem is capturing some delay/shift-invariant structure. In this study, we consider a low-rank model in an embedded space of a tensor. For this purpose, we extend a delay embedding for a time series to a "multi-way delay-embedding transform" for a tensor, which takes a given incomplete tensor as the input and outputs a higher-order incomplete Hankel tensor. The higher-order tensor is then recovered by Tucker-based low-rank tensor factorization. Finally, an estimated tensor can be obtained by using the inverse multi-way delay embedding transform of the recovered higher-order tensor. Our experiments showed that the proposed method successfully recovered missing slices for some color images and functional magnetic resonance images.

Tatsuya Yokota, Burak Erem, Seyhmus Guler, Simon K. Warfield, Hidekata Hontani• 2018

Related benchmarks

TaskDatasetResultRank
Temperature Field PredictionPacific Ocean sea surface temperature dataset (Jan 1970–Dec 1974)
MAE1.44
16
Multidimensional Time Series RecoveryAbilene Pattern-1, 60% missing
MAE0.64
7
Multidimensional Time Series RecoveryAbilene Pattern-1, 80% missing
MAE0.68
7
Multidimensional Time Series RecoveryAbilene Pattern-1, 20% missing
MAE1.04
7
Multidimensional Time Series RecoveryAbilene Pattern-1, 40% missing
MAE0.77
7
Multidimensional Time Series RecoveryAbilene Pattern-2, 30% missing
MAE1.77
7
Multidimensional Time Series RecoveryAbilene Pattern-3, 50% missing
MAE1.36
7
Urban Traffic EstimationNYC-yellow Pattern-1 (first 60 days of 2021)
MAE3.73
7
Multidimensional Time Series RecoveryAbilene Pattern-2 70% missing
MAE1.18
7
Urban Traffic EstimationNYC-yellow Pattern-3 (first 60 days of 2021)
MAE4.65
7
Showing 10 of 11 rows

Other info

Follow for update