Tuesday, March 10, 2026 10:00 am
-
11:00 am
EDT (GMT -04:00)
Rebecca Killick
Lancaster University
Room: M3 3127
Forecasting Nonstationary Processes: What can be done?
The fundamentals of nonstationary forecasting have been considered by several authors, but mostly from a theoretical standpoint. A handful of papers provide an implementable forecast but often rely on user-chosen settings for some parameters that are not easily interpreted in the context of the data at hand. This talk presents a fully implemented and automated approach to forecasting locally stationary time series alongside motivating applications from economics and environmental science.