缺失值归算是如何与批量效应混的,以及您可以做些什么.
Wilson Wen Bin Goh1, Harvard Wai Hann Hui2, Limsoon Wong3
1Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore; School of Biological Sciences, Nanyang Technological University, Singapore; Center for Biomedical Informatics, Nanyang Technological University, Singapore.
批量效应校正和缺失值赋值是相互依赖的数据处理步骤. 了解它们的相互作用可以提高数据质量,并减少先进建模中的错误发现.
科学领域:
- 数据科学数据科学数据科学
- 生物信息学是一种生物信息学.
- 计算生物学 计算生物学
背景情况:
- 数据处理管道依赖于连续的步骤,而上游流程会影响下游流程.
- 批量效应校正 (BEC) 和缺失值归算 (MVI) 对数据在先进建模中的适用性和减少错误发现至关重要.
- BEC和MVI之间的相互作用没有得到广泛的研究,但基本上是相互依赖的.
研究的目的:
- 阐明BEC和MVI的相互联系和相互依赖.
- 展示批量敏感化如何提高MVI质量.
- 引入批量效应相关缺失值 (BEAMs) 的概念,并讨论批量-类不平衡的缓解策略.
主要方法:
- 关于BEC-MVI相互作用的概念讨论.
- 展示批量敏感化对MVI的影响.
- 探索机器学习技术,以解决批量-类不平衡问题.
主要成果:
- 批量敏感化提高了缺失值赋值的质量.
- 对缺失的核算可以在纠正过程中提高批量效应估计.
- 与批量效应相关的缺失值 (BEAM) 被确定为一个关键的交互领域.
- 机器学习概念可以减轻批量类不平衡问题.
结论:
- BEC和MVI是相互依赖的,它们的相互作用显著影响数据质量.
- 批量敏感化提供了一种改善MVI的方法,同时考虑缺失辅助BEC.
- 解决BEAM和批量类不平衡对于强大的数据处理管道至关重要.
更多相关视频
07:34Probing the Limits of Egg Recognition Using Egg Rejection Experiments Along Phenotypic Gradients
Published on: August 22, 2018
06:55Inverse Probability of Treatment Weighting Propensity Score using the Military Health System Data Repository and National Death Index
Published on: January 8, 2020
相关概念视频
Parameters Affecting Nonlinear Elimination: Zero-Order Input, First-Order Absorption and Two-Compartment Model
When a drug is administered through a constant intravenous infusion and eliminated via nonlinear pharmacokinetics, it follows zero-order input. For example, oral drugs undergo first-order absorption upon administration and are eliminated through nonlinear pharmacokinetics.
In the case of subcutaneously administered drugs,...
Strategies for Assessing and Addressing Confounding
Confounding can be addressed at both the design phase of a study and through analytical methods after data...
Mechanistic Models: Compartment Models in Individual and Population Analysis
Confounding in Epidemiological Studies
Estimating Population Mean with Unknown Standard Deviation
William S. Gosset (1876–1937) of the...
One-Compartment Open Model: Wagner-Nelson and Loo Riegelman Method for ka Estimation
On...
