期刊
EUROPEAN PHYSICAL JOURNAL C
卷 81, 期 7, 页码 -出版社
SPRINGER
DOI: 10.1140/epjc/s10052-021-09469-y
关键词
-
资金
- US Department of Energy, Office of Science, Office of High Energy Physics [DE-SC0007890]
- U.S. Department of Energy (DOE) [DE-SC0007890] Funding Source: U.S. Department of Energy (DOE)
In this study, the use of Mixture Density Networks (MDNs) for parameter estimation in situations where training data is only available for discrete values of a continuous parameter is demonstrated. The origins of biases are discussed, and corrective methods for each issue are proposed to address performance-limiting problems that may arise.
Mixture density networks (MDNs) can be used to generate posterior density functions of model parameters theta given a set of observables x. In some applications, training data are available only for discrete values of a continuous parameter theta. 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.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据