Research map / Learning theory and algorithms / Matrix and tensor methods
Orthogonalized ALS: A Theoretically Principled Tensor Decomposition Algorithm for Practical Use
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Add periodic orthogonalization to ALS | Word co-occurrence tensor embeddings, w=3 · WordSim | correlation: 0.47 -> 0.57 | Helped |
| Add periodic orthogonalization to ALS | Word co-occurrence tensor embeddings, w=3 · MEN | correlation: 0.51 -> 0.59 | Helped |
| Add periodic orthogonalization to ALS | Word co-occurrence tensor embeddings, w=3 · Mixed analogies | accuracy: 30.22% -> 38.44% | Helped |
| Add periodic orthogonalization to ALS | Word co-occurrence tensor embeddings, w=3 · Syntactic analogies | accuracy: 32.01% -> 42.40% | Helped |
| Add periodic orthogonalization to ALS | Word co-occurrence tensor embeddings, w=5 · WordSim | correlation: 0.44 -> 0.56 | Helped |
| Add periodic orthogonalization to ALS | Word co-occurrence tensor embeddings, w=5 · MEN | correlation: 0.51 -> 0.60 | Helped |
| Add periodic orthogonalization to ALS | Word co-occurrence tensor embeddings, w=5 · Mixed analogies | accuracy: 37.46% -> 45.87% | Helped |
| Add periodic orthogonalization to ALS | Word co-occurrence tensor embeddings, w=5 · Syntactic analogies | accuracy: 37.27% -> 47.13% | Helped |
| Use tensor embeddings from Orth-ALS instead of matrix SVD embeddings, w=3 | Word embeddings · WordSim | correlation: 0.61 -> 0.57 | Hurt |
| Use tensor embeddings from Orth-ALS instead of matrix SVD embeddings, w=3 | Word embeddings · MEN | correlation: 0.69 -> 0.59 | Hurt |
| Use tensor embeddings from Orth-ALS instead of matrix SVD embeddings, w=5 | Word embeddings · Mixed analogies | accuracy: 54.29% -> 45.87% | Hurt |
| Use tensor embeddings from Orth-ALS instead of matrix SVD embeddings, w=5 | Word embeddings · Syntactic analogies | accuracy: 62.20% -> 47.13% | Hurt |
Closest papers by meaning
- Fused Orthogonal Alternating Least Squares for Tensor Clustering (NeurIPS 2022 · similarity 0.91)
- More Efficient Sampling for Tensor Decomposition With Worst-Case Guarantees (ICML 2022 · similarity 0.89)
- Fast and accurate randomized algorithms for low-rank tensor decompositions (NeurIPS 2021 · similarity 0.89)
- Efficient Leverage Score Sampling for Tensor Train Decomposition (NeurIPS 2024 · similarity 0.88)
- SPALS: Fast Alternating Least Squares via Implicit Leverage Scores Sampling (NeurIPS 2016 · similarity 0.88)
- Performance Gaps in Multi-view Clustering under the Nested Matrix-Tensor Model (ICLR 2024 · similarity 0.86)
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.