Predictive Density Combination Using Bayesian Machine Learning

B-Tier
Journal: International Economic Review
Year: 2025
Volume: 66
Issue: 3
Pages: 1287-1315

Authors (5)

Tony Chernis (Bank of Canada) Niko Hauzenberger (not in RePEc) Florian Huber (not in RePEc) Gary Koop (not in RePEc) James Mitchell (not in RePEc)

Score contribution per author:

0.404 = (α=2.02 / 5 authors) × 1.0x B-tier

α: calibrated so average coauthorship-adjusted count equals average raw count

Abstract

Based on agent opinion analysis theory, Bayesian predictive synthesis (BPS) is a framework for combining predictive distributions in the face of model uncertainty. In this article, we generalize existing parametric implementations of BPS by showing how to combine competing probabilistic forecasts using interpretable Bayesian tree‐based machine learning methods. We demonstrate the advantages of our approach—in terms of improved forecast accuracy and interpretability—via two macroeconomic forecasting applications. The first uses density forecasts for GDP growth from the euro area's Survey of Professional Forecasters. The second combines density forecasts of U.S. inflation produced by many simple regression models.

Technical Details

RePEc Handle
repec:wly:iecrev:v:66:y:2025:i:3:p:1287-1315
Journal Field
General
Author Count
5
Added to Database
2026-01-25