Research map / Learning theory and algorithms / Matrix and tensor methods
Efficient Low Rank Tensor Ring Completion
Wenqi Wang, Vaneet Aggarwal, Shuchin Aeron
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use tensor ring (TR-ALS) instead of tensor train (TT-ALS) | TR-ALS · YaleFace B completion (10% observed) | recovery error (%): 25.55% -> 16.25% | Improved |
| use TR-ALS instead of TT-ALS | TR-ALS · Gun shot video completion (10% observed) | recovery error (%): 14.83% -> 6.25% | Improved |
| use TR-ALS instead of TT-ALS at rank 18 | TR-ALS · Einstein image completion (10% observed) | recovery error (%): 20.70% -> 10.83% | Improved |
| use TR-ALS instead of TT-ALS, any rank | TR-ALS · Einstein image completion (5/10/20/30% observed) | recovery error | Improved |
| reshape YaleFace tensor from 4-order to 8-order (ablation) | TR-ALS · YaleFace B completion (10% observed) | recovery error (%): 21.48% -> 16.25% | Improved |
| reshape further from 8-order to 11-order (ablation) | TR-ALS · YaleFace B completion (10% observed) | recovery error (%): 16.25% -> 16.34% | Hurt |
| increase TR-rank beyond threshold (ablation) | TR-ALS · Einstein image completion | recovery error: 10.83% -> 14.55% | Hurt |
| increase observation ratio (ablation) | TR-ALS · Synthetic 20x20x20x20 TR-rank 8 tensor | convergence speed | Improved |
| use TR-ALS instead of TT-ALS (rank 8 and 16) and SiLRTC | TR-ALS · Synthetic 20x20x20x20 TR-rank 8 tensor | recovery error: 10^-10 for observation ratio larger than 14% | Improved |
| double TT-rank from [8,8,8] to [16,16,16] (ablation) | TT-ALS · Synthetic 20x20x20x20 TR-rank 8 tensor | recovery error | No clear effect |
Closest papers by meaning
- Tensor Ring Decomposition with Rank Minimization on Latent Space: An Efficient Approach for Tensor Completion (AAAI 2019 · similarity 0.92)
- Low-Rank Tensor Completion With a New Tensor Nuclear Norm Induced by Invertible Linear Transforms (CVPR 2019 · similarity 0.91)
- Low-Rank Tensor Completion: A Pseudo-Bayesian Learning Approach (ICCV 2017 · similarity 0.90)
- Tensor Completion Made Practical (NeurIPS 2020 · similarity 0.90)
- Low-Rank Tensor Completion by Approximating the Tensor Average Rank (ICCV 2021 · similarity 0.90)
- Fast Tensor Completion via Approximate Richardson Iteration (ICML 2025 · similarity 0.90)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
