I would also add causal inference (which isn't exactly "new" but i suppose it's still a research field very opened, as a lot of researchers are on it), and the quantile approach for conditional models (the latter may be biased a little bit since it is my research of interest)!
If I had to "choose" one research field I'd definetly lean towards one of these two, assuming that I am interested in either the inference or conditional models (and survival analysis) areas.