Modeling the Diversity of Transient Light Curves through PCA

Nancy Grace Roman Space Telescope would discover a large number of transients which exhibits a wide variety of diversity in light curves.   It is important for us to model photometric light curves with a small number of parameters through PCA (Principal Component Analysis).   PCA can be used for simulations,  and it works well with machine learning techniques. For Roman Space Telescope data, we can apply PCA for classifications and light curve fitting.  We demonstrate PCA application on both time series and diversity / classifications, and show only a few parameters are needed to describe the time series and diversity.