Evaluating probabilistic forecasts with scoringRules
read the original abstract
Probabilistic forecasts in the form of probability distributions over future events have become popular in several fields including meteorology, hydrology, economics, and demography. In typical applications, many alternative statistical models and data sources can be used to produce probabilistic forecasts. Hence, evaluating and selecting among competing methods is an important task. The scoringRules package for R provides functionality for comparative evaluation of probabilistic models based on proper scoring rules, covering a wide range of situations in applied work. This paper discusses implementation and usage details, presents case studies from meteorology and economics, and points to the relevant background literature.
This paper has not been read by Pith yet.
Forward citations
Cited by 2 Pith papers
-
Efficient distributional regression trees learning algorithms for calibrated non-parametric probabilistic forecasts
Novel algorithms for efficient learning of distributional regression trees optimized for CRPS and WIS losses via heaps, balanced trees, and Fenwick trees, with competitive performance and conformal prediction applications.
-
Accurate and Reliable Uncertainty Estimates for Deterministic Predictions Extensions to Under and Overpredictions
Extends ACCRUE to learn input-dependent two-piece Gaussian and asymmetric Laplace uncertainty distributions via neural networks for improved probabilistic forecasts from deterministic models.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.