Keywords: Data distribution shift, probabilistic forecasting, adaptive sampling, importance sampling, Bayesian optimization
TL;DR: We introduce a two-step, model-agnostic method with Bayesian optimization to forecast under distribution shift.
Abstract: The world is not static: This causes real-world time series to change over time through external, and potentially disruptive, events such as macroeconomic cycles or the COVID-19 pandemic. We present an adaptive sampling strategy that selects the part of the time series history that is relevant for forecasting. We achieve this by learning a discrete distribution over relevant time steps by Bayesian optimization. We instantiate this idea with a two-step method that is pre-trained with uniform sampling and then training a lightweight adaptive architecture with adaptive sam- pling. We show with synthetic and real-world experiments that this method adapts to distribution shift and significantly reduces the forecasting error of the base model for three out of five datasets.