Causal and generative time series: research map
674 papers on Causal and generative time series in Time series, from ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR (2016–2026), grouped into 2 clusters and 8 approaches. The busiest year so far is 2026.
What it is
Probabilistic and generative models of time series: forecasting with uncertainty, latent dynamics and state-space models, irregularly sampled data, neural differential equations and diffusion models for time series.
Terms you will see on this page:
- Probabilistic forecasting
- predicting a distribution over future values, not just one number.
- Irregular time series
- observations at uneven times, often with gaps.
- Imputation
- filling in missing values.
Start reading with the most cited papers here:
- Time-series Generative Adversarial Networks (NeurIPS 2019 · 511 citations)
- Deep State Space Models for Time Series Forecasting (NeurIPS 2018 · 429 citations)
Within Time series, its share shrank from 40.9% in 2023–24 to 29.1% in 2025–26 (155 → 229 papers at ICML, NeurIPS, ICLR, CVPR and AAAI, the venues with data for all four years).
Explore Causal and generative time series in the interactive map
Working on something in this topic? Describe your idea in scime atlas to see which approach it falls under, the closest papers by meaning and how crowded the spot has become.
What changed recently
Approaches whose share of the topic moved most between 2023–24 and 2025–26 (ICML, NeurIPS, ICLR, CVPR and AAAI):
- Diffusion models for time series rising: 9% → 16% of the topic (14 → 37 papers)
- Irregular series and imputation rising: 9% → 11% of the topic (14 → 26 papers)
- Probabilistic and conformal forecasting rising: 19% → 21% of the topic (29 → 47 papers)
- Robust forecasting and evaluation falling: 17% → 13% of the topic (27 → 30 papers)
- Dynamical system reconstruction falling: 15% → 12% of the topic (23 → 27 papers)
Papers from 2025–2026 that are already cited most:
- Auto-Regressive Moving Diffusion Models for Time Series Forecasting (AAAI 2025 · 22 citations)
- Self-attention-based Diffusion Model for Time-series Imputation in Partial Blackout Scenarios (AAAI 2025 · 9 citations)
- EBS-EKF: Accurate and High Frequency Event-based Star Tracking (CVPR 2025 · 7 citations)
- DualDynamics: Synergizing Implicit and Explicit Methods for Robust Irregular Time Series Analysis (AAAI 2025 · 7 citations)
Approaches and key papers
Probabilistic and latent forecasting · 503 papers
Keywords: conformal, causal discovery, predictions
Approaches in this cluster:
- Probabilistic and conformal forecasting (139 papers)
Produce calibrated probabilistic or conformal forecasts for series, including hierarchical and non-stationary data. - State-space and latent dynamics forecasting (103 papers)
Forecast via deep state-space, latent-state and dynamics-based models with approximation theory. - Dynamical system reconstruction (99 papers)
Learn latent nonlinear or linear dynamical systems from time series data. - Robust forecasting and evaluation (103 papers)
Study robustness, benchmarking pitfalls and task-based optimization in time series forecasting. - Generative temporal point processes (59 papers)
Uses diffusion, intensity-free and meta-learning methods for temporal point processes.
Most cited and most cited since 2024:
- Time-series Generative Adversarial Networks (NeurIPS 2019 · 511 citations)
- Deep State Space Models for Time Series Forecasting (NeurIPS 2018 · 429 citations)
- CUTS+: High-Dimensional Causal Discovery from Irregular Time-Series (AAAI 2024 · 37 citations)
- Root Cause Analysis in Microservice Using Neural Granger Causal Discovery (AAAI 2024 · 33 citations)
Continuous-time and generative models · 171 papers
Keywords: differential, continuous, equations
Approaches in this cluster:
- Irregular series and imputation (61 papers)
Handle missing values and irregular sampling with probabilistic, generative and continuous-time models. - Neural differential equations (53 papers)
Model irregular time series with neural ODEs, SDEs and controlled differential equations. - Diffusion models for time series (57 papers)
Generate and forecast time series using conditional diffusion models.
Most cited and most cited since 2024:
- BRITS: Bidirectional Recurrent Imputation for Time Series (NeurIPS 2018 · 372 citations)
- Liquid Time-constant Networks (AAAI 2021 · 309 citations)
- Latent Diffusion Transformer for Probabilistic Time Series Forecasting (AAAI 2024 · 45 citations)
- GraFITi: Graphs for Forecasting Irregularly Sampled Time Series (AAAI 2024 · 27 citations)
