Introducing Time-series Foundation Model

About this event

Forecasting is central to decisions in demand planning, energy, operations, and finance. Yet traditional forecasting workflows often require careful model selection, feature engineering, and dataset-specific training before they can be useful. Time-series foundation models offer a different starting point: a model pre-trained across a broad collection of time-series patterns that can produce forecasts for an unseen series with little or no task-specific training. Join us for an technical session on AWS Chronos, a family of pretrained time-series forecasting models. We will unpack the core idea of zero-shot forecasting, walk through how historical signals become probabilistic forecasts, and discuss when these models are useful in a real forecasting workflow. Slides for past meetups posted: Github (https://github.com/YanXuHappygela/LLM-reading-group/tree/main) Recordings posted at: YanAITalk (https://www.youtube.com/@yanaitalk/videos) Feel free to reach out if you want to present at upcoming meetups! Note: You must have a Zoom account to login (free account is sufficient). Zoom Link will be posted to the event page one day before the meetup.