Combining probability forecasts

By Roopesh Ranjan, Tilmann Gneiting. 2010

268 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/e12b72f92c772211e88728ee923a3ff204ed790f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Combining forecasts – forty years later (2011, background): “The same result also obtains for combinations of event probability forecasts (Hora, 2004; Ranjan and Gneiting, 2010).”
The Evolution of Forecast Density Combinations in Economics (2018, background): “…and mathematical statistics The meteorology literature has also been at the vanguard of combinations of probability forecasts, with applications from Sanders (1963) through to a range of more recent papers including Gneiting et al. (2005), Sloughter et al. (2010) and Ranjan and Gneiting (2010).”
Combining Predictive Distributions (2011, methodology): “Introduced by Ranjan and Gneiting (2010) in the context of probability forecasts for a binary event, it generalizes readily to the full class FR of the cumulative distribution functions on R. Specifically, the BLP combination formula maps F1, . . . , Fk ∈ FR to Gα,β ∈ FR, where Gα,β(y) = Bα,β (…”
Combining multiple probability predictions using a simple logit model (2014, methodology): “Given that Ranjan and Gneiting (2010) only considered aggregating binary events, it is not clear how the Beta-transformed linear pooling can be generalized to events with more than two possible outcomes.”
Probability Aggregation Methods in Geoscience (2012, methodology): “A similar derivation for the linear opinion pool and its Beta transform can be found Ranjan and Gneiting (2010).”
Forecasting: theory and practice (2020, background): “As the crowd’s diversity increases, the average probability forecast becomes more spread out, or more underconfident (Dawid et al., 1995; Hora, 2004; Ranjan and Gneiting, 2010).”
Bayesian Nonparametric Calibration and Combination of Predictive Distributions (2015, methodology): “We build on the predictive density calibration framework of Ranjan and Gneiting (2010) and Gneiting and Ranjan (2013) and propose finite and infinite beta mixtures as calibration functions.”
The Good Judgment Project: A Large Scale Test of Different Methods of Combining Expert Predictions (2012, background): “…to averaging different sources such as prediction markets as well as individuals [Graefe et al 2011] However, although the vast majority of people aggregating forecasters use a linear combination of the individual probability estimates, theory shows that no such linear combination can be optimal.”