The Development of New Tensor Decomposition Algorithms to Reveal the Connectivity of Cognitive Learning using EEG

开发新的张量分解算法以揭示脑电图认知学习的连通性

• 批准号: 1941577
• 金额: --
• 依托单位: University of Edinburgh
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2017
• 资助国家: 英国
• 起止时间: 2017 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-1941577
• 关键词: Development; New; Tensor; Decomposition; Algorithms

Learning is a fundamental skill to all people, we spent a big portion of our lives learning. But some are better than others, some acquire skills faster and what makes them different from the rest is likely the way their neural networks are structured.Learning takes time and has various stages which occur in various parts of the brain. One potential method to get a better understanding about the dynamic functional connectivity of learning is by applying tensor decomposition algorithms to EEG data recorded during the learning process. This will decompose the brain activity into it's functional components, for which connectivity can be found using coupling between 2 or more components.This topic aims to develop new tensor decomposition algorithms and bring new insight into the connectivity topology of neural networks during cognitive learning.

学习是所有人的基本技能，我们一生中花了很大一部分时间学习。但是有些人比其他人更好，有些人更快地获得技能，使他们与其他人不同的可能是他们的神经网络结构。学习需要时间，并且在大脑的不同部位发生不同的阶段。一个潜在的方法，以获得更好地了解学习的动态功能连接是通过应用张量分解算法，在学习过程中记录的EEG数据。这将把大脑活动分解成它的功能组件，可以通过两个或多个组件之间的耦合来找到这些组件的连接。本主题旨在开发新的张量分解算法，并对认知学习过程中神经网络的连接拓扑结构带来新的见解。