Probability aggregation in time-series: dynamic hierarchical modeling of sparse expert beliefs

DOI10.1214/14-AOAS739zbMATH Open1429.62706arXiv1408.0087OpenAlexW1968179426MaRDI QIDQ400690FDOQ400690

Authors: Ville A. Satopää, Shane T. Jensen, Barbara A. Mellers, Philip E. Tetlock, Lyle H. Ungar 

Publication date: 22 August 2014

Published in: The Annals of Applied Statistics (Search for Journal in Brave)

Abstract: Most subjective probability aggregation procedures use a single probability judgment from each expert, even though it is common for experts studying real problems to update their probability estimates over time. This paper advances into unexplored areas of probability aggregation by considering a dynamic context in which experts can update their beliefs at random intervals. The updates occur very infrequently, resulting in a sparse data set that cannot be modeled by standard time-series procedures. In response to the lack of appropriate methodology, this paper presents a hierarchical model that takes into account the expert's level of self-reported expertise and produces aggregate probabilities that are sharp and well calibrated both in- and out-of-sample. The model is demonstrated on a real-world data set that includes over 2300 experts making multiple probability forecasts over two years on different subsets of 166 international political events.

Full work available at URL: https://arxiv.org/abs/1408.0087

Recommendations

• Combining probability distributions: A critique and an annotated bibliography
• Aggregating disparate estimates of chance
• Combining probability distributions: a critique and an annotated bibliography. With comments, and a rejoinder by the authors
• Publication:3484159
• Aggregating large sets of probabilistic forecasts by weighted coherent adjustment

zbMATH Keywords

time seriesbias estimationcalibrationdynamic linear modelexpert forecasthierarchical modelingprobability aggregationsubjective probability

Mathematics Subject Classification ID

Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10) Applications of statistics to social sciences (62P25)

Cites Work

• Title not available (Why is that?)
• Slice sampling. (With discussions and rejoinder)
• Bayesian data analysis.
• Title not available (Why is that?)
• Combining predictive distributions
• On Gibbs sampling for state space models
• Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images
• Low Moments for Small Samples: A Comparative Study of Order Statistics
• Title not available (Why is that?)
• A weakly informative default prior distribution for logistic and other regression models
• Combining probability distributions: A critique and an annotated bibliography
• Probability aggregation methods in geoscience
• Rejoinder on: Assessing probabilistic forecasts of multivariate quantities, with an application to ensemble predictions of surface winds
• Title not available (Why is that?)
• Probability aggregation in time-series: dynamic hierarchical modeling of sparse expert beliefs
• Two reasons to make aggregated probability forecasts more extreme
• Comments on: Assessing probabilistic forecasts of multivariate quantities, with an application to ensemble predictions of surface winds

Cited In (3)

• Probability aggregation in time-series: dynamic hierarchical modeling of sparse expert beliefs
• Aggregating predictions from experts: a review of statistical methods, experiments, and applications
• Multivariate Bayesian predictive synthesis in macroeconomic forecasting

Uses Software

• BayesDA

This page was built for publication: Probability aggregation in time-series: dynamic hierarchical modeling of sparse expert beliefs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q400690)