Ilmun Kim

New semi-supervised method enhances statistical estimation with unlabeled data.

This article discusses creating reliable hypothesis tests despite data corruption.

Investigating faster methods for two-sample testing using random Fourier features.