Overlapping clustering of time-dependent variables for fMRI data
提出一种处理时间依赖数据的重叠聚类方法，基于静息态fMRI时间序列识别大脑的重叠粗分区，也可用于基因表达等其他多时间序列场景。
Abstract We propose in this paper a framework for performing overlapping clustering in the presence of time dependence, where the main goal is to identify overlapping coarse parcelations of the brain based on resting state fMRI time series. Our procedure is based on the Latent OVErlapping (LOVE) clustering method of Bing et. al (2020) which we extend to weakly dependent time series. Although the method is developed in an fMRI context, it is general enough to be directly applicable to other contexts, such as gene expression data, where one has at disposal multiple time series and is interested in identifying overlapping groups of similar elements.