Research map / Learning theory and algorithms / Matrix and tensor methods
Large-Scale Learning with Fourier Features and Tensor Decompositions
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 |
|---|---|---|---|
| Use a low-rank CP tensor over deterministic Fourier features instead of random Fourier features | T-KRR · Yacht, Energy, Airfoil, Spambase, Adult | MSE or misclassification rate: Yacht: 0.0021±0.0018 -> 0.0009±0.0006; Energy: 0.0275±0.0075 -> 0.0200±0.0066; Airfoil: 0.2180± 0.0336 -> 0.1679±0.0258; | Helped |
| Use a low-rank CP tensor over deterministic Fourier features instead of full KRR | T-KRR · Yacht, Energy, Airfoil, Spambase | MSE or misclassification rate: Yacht: 0.0007±0.0004 -> 0.0009±0.0006; Energy: 0.0200±0.0087 -> 0.0200±0.0066; Airfoil: 0.1587±0.0232 -> 0.1679±0.0258; | Mixed |
| Increase CP-rank from 5 to 10 (ablation) | Airline T-KRR · Airline, N=10000 | MSE: 0.91±0.10 -> 0.89±0.05 | Helped |
| Increase CP-rank from 5 to 15 (ablation) | Airline T-KRR · Airline, N=10000 | MSE: 0.91±0.10 -> 0.90±0.07 | Mixed |
| Increase CP-rank from 5 to 20 (ablation) | Airline T-KRR · Airline, N=10000 | MSE: 0.91±0.10 -> 0.97±0.15 | Hurt |
| Increase CP-rank from 5 to 10 (ablation) | Airline T-KRR · Airline, N=100000 | MSE: 0.82±0.03 -> 0.80±0.05 | Helped |
| Increase CP-rank from 5 to 15 (ablation) | Airline T-KRR · Airline, N=100000 | MSE: 0.82±0.03 -> 0.80±0.04 | Helped |
| Increase CP-rank from 5 to 20 (ablation) | Airline T-KRR · Airline, N=100000 | MSE: 0.82±0.03 -> 0.78±0.04 | Helped |
| Increase CP-rank from 5 to 20 (ablation) | Airline T-KRR · Airline, N=1000000 | MSE: 0.80±0.02 -> 0.77±0.01 | Helped |
| Increase CP-rank from 5 to 20 (ablation) | Airline T-KRR · Airline, N=5929413 | MSE: 0.800±0.008 -> 0.763±0.007 | Helped |
| Use CP-rank 6 instead of the full matrix rank (ablation) | Banana T-KRR · Banana | decision boundary agreement | No clear effect |
Closest papers by meaning
- Towards a Unified Analysis of Random Fourier Features (ICML 2019 · similarity 0.90)
- Nonparametric Decomposition of Sparse Tensors (ICML 2021 · similarity 0.89)
- Sharp Analysis of Random Fourier Features in Classification (AAAI 2022 · similarity 0.89)
- Random features for Grassmannian kernel approximation with bounded rank-one projections (NeurIPS 2026 · similarity 0.87)
- Efficiently Learning Significant Fourier Feature Pairs for Statistical Independence Testing (NeurIPS 2024 · similarity 0.86)
- Generalized Tensor Decomposition for Understanding Multi-Output Regression under Combinatorial Shifts (NeurIPS 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.