作者: Fabian J. Theis , Peter Gruber , Ingo R. Keck , Elmar W. Lang
DOI: 10.1016/J.NEUCOM.2007.06.012
关键词:
摘要: Real-world data sets such as recordings from functional magnetic resonance imaging (fMRI) often possess both spatial and temporal structures. Here, we propose an algorithm including spatiotemporal information into the analysis, reduce problem to joint approximate diagonalization of a set autocorrelation matrices. We demonstrate feasibility by applying it fMRI where previous approaches are outperformed considerably.