 |
| Titel |
scoringRules – A software package for probabilistic model evaluation |
| VerfasserIn |
Sebastian Lerch, Alexander Jordan, Fabian Krüger |
| Konferenz |
EGU General Assembly 2016
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 18 (2016) |
| Datensatznummer |
250122552
|
| Publikation (Nr.) |
 EGU/EGU2016-1613.pdf |
|
|
|
|
|
| Zusammenfassung |
| Models in the geosciences are generally surrounded by uncertainty, and being able to quantify
this uncertainty is key to good decision making. Accordingly, probabilistic forecasts in the
form of predictive distributions have become popular over the last decades. With
the proliferation of probabilistic models arises the need for decision theoretically
principled tools to evaluate the appropriateness of models and forecasts in a generalized
way.
Various scoring rules have been developed over the past decades to address this demand.
Proper scoring rules are functions S(F,y) which evaluate the accuracy of a forecast
distribution F , given that an outcome y was observed. As such, they allow to compare
alternative models, a crucial ability given the variety of theories, data sources and statistical
specifications that is available in many situations.
This poster presents the software package scoringRules for the statistical
programming language R, which contains functions to compute popular scoring rules such as
the continuous ranked probability score for a variety of distributions F that come up in
applied work. Two main classes are parametric distributions like normal, t, or gamma
distributions, and distributions that are not known analytically, but are indirectly described
through a sample of simulation draws. For example, Bayesian forecasts produced via Markov
Chain Monte Carlo take this form. Thereby, the scoringRules package provides a
framework for generalized model evaluation that both includes Bayesian as well as classical
parametric models.
The scoringRules package aims to be a convenient dictionary-like reference for
computing scoring rules. We offer state of the art implementations of several known (but not
routinely applied) formulas, and implement closed-form expressions that were previously
unavailable. Whenever more than one implementation variant exists, we offer statistically
principled default choices. |
|
|
|
|
|
|