Modeling the uncertainty on the covariance matrix for probabilistic forecast reconciliation
Modeling the uncertainty on the covariance matrix for probabilistic forecast reconciliation
In forecast reconciliation, the covariance matrix of the base forecasts errors plays a crucial role. Typically, this matrix is estimated, and then treated as known. In contrast, we propose a Bayesian reconciliation model that explicitly accounts for the uncertainty in the covariance matrix. We choose an Inverse-Wishart prior, which leads to a multivariate-t reconciled predictive distribution and allows a completely analytical derivation. Empirical experiments demonstrate that this approach improves the accuracy of the prediction intervals with respect to MinT, leading to more reliable probabilistic forecasts.
Chiara Carrara、Lorenzo Zambon、Dario Azzimonti、Giorgio Corani
数学
Chiara Carrara,Lorenzo Zambon,Dario Azzimonti,Giorgio Corani.Modeling the uncertainty on the covariance matrix for probabilistic forecast reconciliation[EB/OL].(2025-06-24)[2025-07-16].https://arxiv.org/abs/2506.19554.点此复制
评论