A diversity-based predictive prior added to the latent weight dynamics of a Bayesian forecast combination method (DTVW) improves point and density forecasts in simulations and in oil price and U.S. macro applications, compared with equal weights, BMA, and baseline TVW.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Bayesian Forecast Combination with Predictive Priors via Particle Filtering
A diversity-based predictive prior added to the latent weight dynamics of a Bayesian forecast combination method (DTVW) improves point and density forecasts in simulations and in oil price and U.S. macro applications, compared with equal weights, BMA, and baseline TVW.