Brier scores reflect the accuracy of a probability forecast. The scores range from 0 to 1.0. A perfect prediction = 0 meaning complete or total accuracy.
Brier scores are named for Glenn W. Brier
For an example of Brier scores in research, see the metamemory study by Parlett-Pelleriti et al. (2021).
See also a discussion of Brier scores in the book, Noise.
Reference
Kahneman, D., Sibony, O., & Sunstein, C.R. (2021). Noise: A flaw in human judgment. New York: Hatchette.
Parlett-Pelleriti, C., Lin, G., Jones, M., Linstead, E. & Jaeggi, S. (2019). Exploring Age-Related Metamemory Differences using Modified Brier Scores and Hierarchical Clustering. Open Psychology, 1(1), 215-238. https://doi.org/10.1515/psych-2018-0015
No comments:
Post a Comment