Strictly Proper Scoring Rules

138 citations recorded by the source corpus.

How researchers see this paper

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

DISCO Nets : DISsimilarity COefficients Networks (2016, methodology): “By theorem 5 in Gneiting and Raftery [5], this is the case, for example, if we take as loss function ∆β(y,y′) = ||y − y′||β2 = ( ∑dy i=1 |(yi − y′i|2) β/2 with β ∈ [0, 2] excluding 0 and 2, and use γ = 12 .”
Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles (2016, methodology): “Scoring rules measure the quality of predictive uncertainty (see (Gneiting and Raftery, 2007) for a review).”
The General Projected Normal Distribution of Arbitrary Dimension: Modeling and Bayesian Inference (2017, methodology): “We also randomly selected 295 proteins to set aside, so that n = 15, 000 angle pairs from 1000 proteins were used to fit the mixture model, and m = (15)(295) = 4425 from 295 proteins were set aside for out-of-sample model comparison based on the PLSL.”