Probabilistic reconciliation of count time series

Abstract

Forecast reconciliation is an important research topic. Yet, there is currently neither a formal framework nor a practical method for the probabilistic reconciliation of count time series. This paper proposes a definition of coherency and reconciled probabilistic forecast, which applies to real-valued and count variables, and a novel method for probabilistic reconciliation. It is based on a generalization of Bayes’ rule and can reconcile real-value and count variables. When applied to count variables, it yields a reconciled probability mass function. Our experiments with the temporal reconciliation of count variables show a major forecast improvement compared to the probabilistic Gaussian reconciliation.