Researchers have developed a novel framework called LLM as Forecasting Planner (rc) that integrates large language models (LLMs) with time-series foundation models (TSFMs) for improved text-conditioned forecasting. This training-free approach treats forecasting as a planning problem, where the TSFM generates numerical continuations and the LLM guides the selection and evaluation of these trajectories based on natural language context. Experiments using TSFMs like Chronos and TimesFM, alongside various LLMs, demonstrated consistent performance gains across different model combinations. AI
IMPACT This framework offers a new method for text-conditioned time-series forecasting by leveraging LLMs without retraining, potentially improving predictions in domains sensitive to contextual events.
RANK_REASON The item describes a novel framework and experimental results presented in a research paper. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Hugging Face Daily Papers →
- Chronos
- LLM as Forecasting Planner
- Monte Carlo tree search
- Ranker LLM
- Time-MMD
- Time Series Foundation Models
- TimesFM
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
