Mixture density network estimation of continuous variable maximum likelihood using discrete training samples
| Autor: | Spencer Stubbs, Charles Burton, Peter Onyisi |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
FOS: Computer and information sciences
Computer Science - Machine Learning Physics and Astronomy (miscellaneous) Maximum likelihood FOS: Physical sciences QC770-798 Astrophysics 01 natural sciences High Energy Physics - Experiment Machine Learning (cs.LG) Continuous variable Set (abstract data type) High Energy Physics - Experiment (hep-ex) Nuclear and particle physics. Atomic energy. Radioactivity 0103 physical sciences Mixture distribution 010306 general physics Engineering (miscellaneous) Discrete mathematics Physics Training set 010308 nuclear & particles physics Estimation theory Observable QB460-466 Physics - Data Analysis Statistics and Probability Continuous parameter Data Analysis Statistics and Probability (physics.data-an) |
| Zdroj: | European Physical Journal C: Particles and Fields, Vol 81, Iss 7, Pp 1-9 (2021) European Physical Journal |
| ISSN: | 1434-6052 1434-6044 |
| DOI: | 10.1140/epjc/s10052-021-09469-y |
| Popis: | Mixture density networks (MDNs) can be used to generate posterior density functions of model parameters $$\varvec{\theta }$$ θ given a set of observables $${\mathbf {x}}$$ x . In some applications, training data are available only for discrete values of a continuous parameter $$\varvec{\theta }$$ θ . In such situations, a number of performance-limiting issues arise which can result in biased estimates. We demonstrate the usage of MDNs for parameter estimation, discuss the origins of the biases, and propose a corrective method for each issue. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink