I was in college during the early late 90s, early 2000s, that was a winter time for ML.Besides the perennial AI, the popular term was "soft computing" to include not only neural networks, but also stuff like fuzzy logic, evolutionary algorithms and simulated annealing. I found it all the approaches so fascinating. For better or for worse the DL explosion totally eclipsed the rest of approaches. Does anybody know of recent relevant work on those fields?