Event Details

CMAF Webinair

In this talk, we will explore the topic of rolling origin forecast instability: what it is, why it matters, and how to address it effectively.

Rolling origin forecast instability refers to the variability in forecasts for a specific time period caused by updates made as new data becomes available over time. While these updates typically improve forecast quality because they are based on a shorter forecast horizon, they can also intro-duce instability, possibly leading to costly changes to plans formulated based on the forecasts. Fur-thermore, forecast instability may erode trust in the forecasting system, potentially prompting un-warranted judgmental adjustments by users.

We will discuss existing methodologies to improve forecast stability without compromising forecast quality, including approaches for both point and probabilistic forecasting. Join us to gain insights into how to balance forecast quality and stability to build more stable and trusted forecasting sys-tems.