Research map / Time series / Causal and generative time series
GluonTS: Probabilistic and Neural Time Series Modeling in Python
Alexander Alexandrov, Konstantinos Benidis, Michael Bohlke-Schneider, Valentin Flunkert, Jan Gasthaus, Tim Januschowski, Danielle C. Maddix, Syama Rangapuram, David Salinas, Jasper Schulz, Lorenzo Stella, Ali Caner Türkmen and 1 more
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 neural global model DeepAR instead of classical Auto-ETS | DeepAR · electricity | mean quantile loss: 0.121 -> 0.050 | Improved |
| use neural global model DeepAR instead of classical Auto-ETS | DeepAR · solar-energy | mean quantile loss: 1.778 -> 0.398 | Improved |
| use neural global model DeepAR instead of classical Auto-ETS | DeepAR · traffic | mean quantile loss: 0.373 -> 0.126 | Improved |
| use neural global model DeepAR instead of classical Auto-ETS | DeepAR · m4-monthly | mean quantile loss: 0.099 -> 0.115 | Hurt |
| use neural global model DeepAR instead of classical Auto-ETS | DeepAR · exchange-rate | mean quantile loss: 0.008 -> 0.023 | Hurt |
| use Transformer instead of Seasonal Naive | Transformer · m4-hourly | mean quantile loss: 0.048 -> 0.035 | Improved |
| use CNN-QR instead of DeepAR | CNN-QR · exchange-rate | mean quantile loss: 0.023 -> 0.016 | Improved |
| use CNN-QR instead of DeepAR | CNN-QR · solar-energy | mean quantile loss: 0.398 -> 0.551 | Hurt |
| use Transformer instead of DeepAR | Transformer · m4-weekly | mean quantile loss: 0.048 -> 0.083 | Hurt |
| use DeepAR instead of Seasonal Naive | DeepAR · traffic | mean quantile loss: 0.251 -> 0.126 | Improved |
Closest papers by meaning
- GlucoFM-Bench: Benchmarking Time-Series Foundation Models for Blood Glucose Forecasting (NeurIPS 2026 · similarity 0.84)
- Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Experts (ICLR 2025 · similarity 0.84)
- Timer: Generative Pre-trained Transformers Are Large Time Series Models (ICML 2024 · similarity 0.83)
- Time-LLM: Time Series Forecasting by Reprogramming Large Language Models (ICLR 2024 · similarity 0.83)
- LightGTS: A Lightweight General Time Series Forecasting Model (ICML 2025 · similarity 0.83)
- SynTSBench: Rethinking Temporal Pattern Learning in Deep Learning Models for Time Series (NeurIPS 2025 · similarity 0.82)
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.
