Research map / Time series / Causal and generative time series
Learning low-dimensional state embeddings and metastable clusters from time series data
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 |
|---|---|---|---|
| reshape kernel mean embedding using the low-rank SVD structure instead of plain KME | KME transition kernel estimator · four-well diffusion, r=4, N=2000 random Fourier features | estimation error | Helped |
| embed DQN hidden-layer states with the proposed embedding before t-SNE | DQN last-hidden-layer time series, Demon Attack · Atari Demon Attack, 47936 x 512 series | cluster structure in t-SNE (qualitative) | Helped |
| increase time interval tau from 0.1 to 10 (ablation) | state embedding and clustering · diffusion process, n=10^6, r=4 | dynamics scale captured (qualitative) | Mixed |
Closest papers by meaning
- Identifying latent state transitions in non-linear dynamical systems (ICLR 2025 · similarity 0.87)
- Latent Representation and Simulation of Markov Processes via Time-Lagged Information Bottleneck (ICLR 2024 · similarity 0.87)
- STEP: Learning STructured Embeddings for Progressive Time Series (NeurIPS 2026 · similarity 0.87)
- Sequence Modeling with Spectral Mean Flows (NeurIPS 2025 · similarity 0.85)
- Structured Sparse Transition Matrices to Enable State Tracking in State-Space Models (NeurIPS 2025 · similarity 0.85)
- Learning invariant representations of time-homogeneous stochastic dynamical systems (ICLR 2024 · similarity 0.85)
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.