Forecast Combination and Interpretability Using Random Subspace
Boris Kozyrev
IWH Discussion Papers,
Nr. 21,
2024
Abstract
This paper investigates forecast aggregation via the random subspace regressions method (RSM) and explores the potential link between RSM and the Shapley value decomposition (SVD) using the US GDP growth rates. This technique combination enables handling high-dimensional data and reveals the relative importance of each individual forecast. First, it is possible to enhance forecasting performance in certain practical instances by randomly selecting smaller subsets of individual forecasts and obtaining a new set of predictions based on a regression-based weighting scheme. The optimal value of selected individual forecasts is also empirically studied. Then, a connection between RSM and SVD is proposed, enabling the examination of each individual forecast’s contribution to the final prediction, even when there is a large number of forecasts. This approach is model-agnostic (can be applied to any set of predictions) and facilitates understanding of how the aggregated prediction is obtained based on individual forecasts, which is crucial for decision-makers.
Artikel Lesen
Department Profiles
Research Profiles of the IWH Departments All doctoral students are allocated to one of the four research departments (Financial Markets – Laws, Regulations and Factor Markets –…
Zur Seite
Alumni
Alumni IWH provides guidance and support in job placement after graduation, including letters of recommendation and career advice. Graduates have found placements in academia…
Zur Seite
Placement News
Placement News IWH-DPE graduate placements 2024: St Andrews and the Bank of England Eleonora Sfrappini from the Financial Markets department and Tommaso Bighelli from the…
Zur Seite
Past Events
Past Events 14. CompNet Annual Conference (Vilnius, 25-26 September 2025) The 14th CompNet Annual Conference, co-hosted with the Bank of Lithuania, took place on 25–26 September…
Zur Seite
Trade and Competitiveness
Trade and Competitiveness The investigation delves into the mechanisms through which productivity diffuses within Global Value Chains (GVC), assessing how national firms respond…
Zur Seite
Compnet Training Program
CompNet Training Program Structure The course is made for autonomous online learning. It is structured in three modules : Beginners, Intermediate and Advanced. Each of them…
Zur Seite
13 CompNet Annual Conference
13th CompNet Annual Conference The 13th Annual Conference in Valletta, hosted by the Central Bank of Malta, was a resounding success. We extend our heartfelt thanks to everyone…
Zur Seite
6th CompNet Annual Conference
Innovation, firm size, productivity and imbalances in the age of de-globalization 6 th CompNet Annual Conference, June 29-30, 2017, European Commission, Brussels, Belgium As the…
Zur Seite
Country Visits 2023
Country Visits In a forthcoming expedition from November 29th to December 1st, 2023, CompNet is set to embark on an insightful journey in Helsinki, Finland. On the agenda are…
Zur Seite
