Learning and Applying a Function Over Distributions
Autor: | Shiyuan Zhao, Glenn Healey |
---|---|
Jazyk: | angličtina |
Rok vydání: | 2020 |
Předmět: |
General Computer Science
02 engineering and technology Computer Science::Computational Geometry Baseball 01 natural sciences Cross-validation 010104 statistics & probability 0202 electrical engineering electronic engineering information engineering General Materials Science nonparametric 0101 mathematics earth mover’s distance Mathematics function over distributions General Engineering Nonparametric statistics 020206 networking & telecommunications Function (mathematics) Distribution (mathematics) machine learning Kernel (statistics) kernel regression Metric (mathematics) Kernel smoother Kernel regression lcsh:Electrical engineering. Electronics. Nuclear engineering Algorithm lcsh:TK1-9971 |
Zdroj: | IEEE Access, Vol 8, Pp 172196-172203 (2020) |
ISSN: | 2169-3536 |
Popis: | We present a method for learning a function over distributions. The method is based on generalizing nonparametric kernel regression by using the earth mover’s distance as a metric for distribution space. The technique is applied to the problem of learning the dependence of pitcher performance in baseball on multidimensional pitch distributions that are controlled by the pitcher. The distributions are derived from sensor measurements that capture the physical properties of each pitch. Finding this dependence allows the recovery of optimal pitch frequencies for individual pitchers. This application is amenable to the use of signatures to represent the distributions and a whitening step is employed to account for the correlations and variances of the pitch variables. Cross validation is used to optimize the kernel smoothing parameter. A set of experiments demonstrates that the new method accurately predicts changes in pitcher performance in response to changes in pitch distribution and also outperforms an existing technique for this application. |
Databáze: | OpenAIRE |
Externí odkaz: |