Research map / Learning theory and algorithms / Matrix and tensor methods
Learning brain regions via large-scale online structured sparse dictionary learning
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 |
|---|---|---|---|
| adds Laplacian (Sobolev) spatial penalty to sparse online dictionary learning (Smooth-SODL) | SODL dictionary learning · HCP task fMRI, 500 subjects | dictionary stability (mean correlation across splits): 0.1 -> 0.4 | Helped |
| adds Laplacian spatial penalty to SODL | SODL dictionary learning · HCP learning curve, few subjects (17 pooled) | mean test explained variance: 2% -> 31% | Helped |
| adds Laplacian spatial penalty to SODL | SODL dictionary learning · HCP learning curve, 241 subjects pooled | mean test explained variance: 49% -> 55% | Helped |
| adds Laplacian spatial penalty to SODL | SODL dictionary learning · HCP task fMRI, 500 subjects | running time: 110s -> 340s | Hurt |
| CanICA (ICA, no sparsity) vs Smooth-SODL on explained variance | CanICA · HCP task fMRI, 500 subjects | explained variance (test): 60% vs 55% | Mixed |
| Laplacian penalty vs CanICA on behavioral prediction | Smooth-SODL · HCP behavioral prediction (R2 on test) | R2 score | Mixed |
| hard-thresholding CanICA components | CanICA · HCP task fMRI, 500 subjects | dictionary stability: 0.1 -> 0.2 | Helped |
| very large gamma (1e8) on Laplacian penalty (ablation) | Smooth-SODL · HCP hyper-parameter study | explained variance | Hurt |
Closest papers by meaning
- Learning Macroscopic Brain Connectomes via Group-Sparse Factorization (NeurIPS 2019 · similarity 0.88)
- Dictionary Learning for Massive Matrix Factorization (ICML 2016 · similarity 0.87)
- Dictionary Learning Based on Sparse Distribution Tomography (ICML 2017 · similarity 0.86)
- Learning sparse codes from compressed representations with biologically plausible local wiring constraints (NeurIPS 2020 · similarity 0.86)
- Sparse components distinguish visual pathways & their alignment to neural networks (ICLR 2025 · similarity 0.84)
- On the Sparsity-Storage-Accuracy Tradeoff in Parsimoniously Activated Dictionary Learning (NeurIPS 2026 · similarity 0.84)
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.