Abstract: The overall goal of the paper is to develop a deep kernel principal component analysis (KPCA) for time-dependent data that are nonlinearly distributed in high dimensions. Instead of ...