block-term-decomposition

[IEEE TKDE 2023] A list of up-to-date papers on streaming tensor decomposition, tensor tracking, dynamic tensor analysis

Interpretable block-term tensor regression (BTTR) for identifying atrial fibrillation in sinus-rhythm ECG on PTB-XL, benchmarked against clinical, CNN, CP, and foundation-model baselines.

Interpretable Block-Term Tensor Network (BTTN) for predicting future-onset (incident) atrial fibrillation from a single sinus-rhythm 12-lead ECG on MIMIC-IV-ECG: the glass-box (time x lead) factor parameters ARE the explanation, with a measured-faithfulness framework, at parity with a CNN. Patient-grouped CV, patient-bootstrap CIs.

Lead-field-constrained block-term decomposition of evoked EEG: solves the EEG inverse problem and decomposes the signal jointly, returning anatomically-grounded spatiotemporal components (a focal cortical source map, a time course, and a trial loading per block).

[ICASSP 2026] Re-LL1: An Effective Regularized ( L , L , 1 ) -Tensor Decomposition Method For Video Background Modeling and Foreground Separation. In IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

Block-Term Tensor Regression (BTTR): predicts a tensor Y from a tensor X by deflation, with automatic component extraction (ACE/ACCoS). Code for Faes et al., IEEE TNNLS 2022.

Verification code for block-term operator theory: why rank-(L,L,1) block-term neural operators generalise better than CP, Tucker and TT at matched capacity. Four scripts, no data or GPU needed.

[SSP 2023] A Novel Tensor Tracking Algorithm For Block-Term Decomposition of Streaming Tensors. In Proc. 22nd IEEE Statistical Signal Processing Workshop (SSP)