The amount of cosmological information we will be able to retrieve from future large scale surveys will be bounded by our ability in photometrically identifying type Ia Sne. This derives from the fact that it is impossible to provide spectroscopic follow up for all Sne discovered by ongoing searches. In this talk, I will present a glimpse of different efforts applied to this problem and take a closer look in the use of kernel Principal Component Analysis as a machine learning technique able to provide high purity results in the final Ia sample.

Download the video:
09-19-2012_Kernel_PCA.avi (55min, 125MB)