Forecast with Forecasts: Diversity Matters
Yanfei Kang, Wei Cao, Fotios Petropoulos, Feng Li

TL;DR
This paper proposes a forecast combination method that leverages forecast diversity and out-of-sample forecasts to improve accuracy, simplifying modeling while achieving superior results in point and interval forecasts.
Contribution
It introduces a novel approach focusing on forecast outputs rather than historical data for feature extraction, enhancing forecast combination performance.
Findings
Outperforms traditional methods in diverse datasets
Simplifies the forecast modeling process
Achieves superior point and interval forecast accuracy
Abstract
Forecast combinations have been widely applied in the last few decades to improve forecasting. Estimating optimal weights that can outperform simple averages is not always an easy task. In recent years, the idea of using time series features for forecast combination has flourished. Although this idea has been proved to be beneficial in several forecasting competitions, it may not be practical in many situations. For example, the task of selecting appropriate features to build forecasting models is often challenging. Even if there was an acceptable way to define the features, existing features are estimated based on the historical patterns, which are likely to change in the future. Other times, the estimation of the features is infeasible due to limited historical data. In this work, we suggest a change of focus from the historical data to the produced forecasts to extract features. We…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
