Local prediction pools

We propose local prediction pools as a method for combining the predictive distributions of a set of experts conditional on a set of variables believed to be related to the predictive accuracy of the experts. This is done in a two-step process where we first estimate the conditional predictive accuracy of each expert given a vector of covariates-or pooling variables-and then combine the predictive distributions of the experts conditional on this local predictive accuracy. To estimate the local predictive accuracy of each expert, we introduce the simple, fast, and interpretable caliper method. Expert pooling weights from the local prediction pool approaches the equal weight solution whenever there is little data on local predictive performance, making the pools robust and adaptive. We also propose a local version of the widely used optimal prediction pools. Local prediction pools are shown to outperform the widely used optimal linear pools in a macroeconomic forecasting evaluation and in predicting daily bike usage for a bike rental company.

Automated summary

We propose a method called local prediction pools, which combines the predictive distributions of a group of experts based on variables believed to be related to their predictive accuracy. This two-step process involves estimating the conditional predictive accuracy of each expert using a set of covariates and then combining their predictive distributions based on this local predictive accuracy. The caliper method is introduced as a simple, fast, and interpretable approach to estimate the local predictive accuracy of each expert. The results show that local prediction pools outperform the widely used optimal linear pools in macroeconomic forecasting and predicting daily bike usage for a bike rental company.