Research map / Probabilistic and scientific modeling / Molecules and proteins
Learning Low-Dimensional Temporal Representations
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 |
|---|---|---|---|
| learn latent alignments jointly with the subspace instead of fixing alignments from the original space (ini-LT-LDA) (ablation) | LT-LDA · Olympic Sports | accuracy and MAP with SVM | Helped |
| increase template length L beyond about 9 | LT-LDA · MSR Action3D | multi-class performance | Hurt |
| use too short template length L | LT-LDA · MSR Action3D | multi-class performance | Hurt |
| set L within 6 to 9 | LT-LDA · MSR Action3D | multi-class performance | Helped |
| use too large warping factor a | LT-LDA · MSR Action3D | performance | Hurt |
| use too small warping factor a | LT-LDA · MSR Action3D | performance | Hurt |
| replace LSDA with LT-LDA under the DTW classifier | LSDA · ChaLearn and Action3D | accuracy: margin of more than 10% | Helped |
| replace LDA and kernel LDA (treating frames as i.i.d.) with LT-LDA | LDA, kLDA · ChaLearn and Action3D | accuracy and MAP | Helped |
| use LT-LDA with under 15 dimensions instead of original features with hundreds of dimensions | original features · ChaLearn and Action3D, HMM and DTW | accuracy | Helped |
| use LT-LDA with 15 or 25 dimensions instead of original features | original features · ChaLearn and Action3D, SVM rank pooling | MAP | No clear effect |
| replace PCA with LT-LDA | PCA · Olympic Sports | accuracy and MAP | Helped |
| preserve more than 19 dimensions with LT-LDA (versus LDA limited to C-1=19) | LDA · Olympic Sports | accuracy and MAP | Helped |
Closest papers by meaning
- Learning Without Augmenting: Unsupervised Time Series Representation Learning via Frame Projections (NeurIPS 2025 · similarity 0.86)
- Learning Low-dimensional Latent Dynamics from High-dimensional Observations: Non-asymptotics and Lower Bounds (ICML 2024 · similarity 0.85)
- Kernel Representation Learning for Time Sequence: Algorithm, Theory, and Applications (AAAI 2025 · similarity 0.85)
- Optimal Representation Size: High-Dimensional Analysis of Pretraining and Linear Probing (NeurIPS 2026 · similarity 0.83)
- Integrating Sequence and Image Modeling in Irregular Medical Time Series Through Self-Supervised Learning (AAAI 2025 · similarity 0.83)
- Temporal Streaming Batch Principal Component Analysis for Time Series Classification (Student Abstract) (AAAI 2025 · similarity 0.83)
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.