Research map / Learning theory and algorithms / Matrix and tensor methods
Efficient frequent directions algorithms for approximate decomposition of matrices and higher-order tensors
Maolin Che, Yimin Wei, Hong Yan
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 SpFD sketch instead of plain FD inside Algorithm 5 | Algorithm 5 · Yale | running time (s): 16.17 -> 1.60 | Improved |
| use SpFD sketch instead of plain FD inside Algorithm 5 | Algorithm 5 · COIL-100 | running time (s): 18.11 -> 2.55 | Improved |
| use SpFD sketch instead of plain FD inside Algorithm 5 | Algorithm 5 · Washington DC | running time (s): 3.43 -> 0.56 | Improved |
| use Gaussian-SpFD instead of FD inside Algorithm 5 | Algorithm 5 · Yale | relative error (RE): 2.17e-1 -> 2.26e-1 | Hurt |
| use Dct-SpFD instead of FD inside Algorithm 5 | Algorithm 5 · COIL-100 | relative error (RE): 4.53e-1 -> 4.57e-1 | Hurt |
| use Algorithm 5 with SpFD instead of Tucker-ALS | Tucker approximation · Yale | relative error (RE): 3.45e-1 -> 2.17e-1 | Improved |
| use Algorithm 5 with SpFD instead of Tucker-ALS | Tucker approximation · COIL-100 | running time (s): 6.43 -> 2.55 | Improved |
| use Algorithm 5 with SpFD instead of Tucker-ALS | Tucker approximation · Washington DC | relative error (RE): 2.30e-1 -> 2.44e-1 | Mixed |
| use Algorithm 5 with SpFD instead of Tucker-TS | Tucker approximation · Yale | running time (s): 501.03 -> 1.60 | Improved |
| use Algorithm 5 with SpFD instead of Sketch-Tucker-ALS | Tucker approximation · COIL-100 | relative error (RE): 4.58e-1 -> 4.53e-1 | No clear effect |
Closest papers by meaning
- A Fast Frequent Directions Algorithm for Low Rank Approximation (TPAMI 2018 · similarity 0.90)
- Fitting Low-Rank Tensors in Constant Time (NeurIPS 2017 · similarity 0.89)
- Fast Tucker Rank Reduction for Non-Negative Tensors Using Mean-Field Approximation (NeurIPS 2021 · similarity 0.88)
- Low-Rank Tucker Decomposition of Large Tensors Using TensorSketch (NeurIPS 2018 · similarity 0.88)
- Nonparametric Decomposition of Sparse Tensors (ICML 2021 · similarity 0.88)
- Fast Tensor Completion via Approximate Richardson Iteration (ICML 2025 · similarity 0.87)
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.
