3 ms·AlphaGo relied heavily on (supervised) pretraining, and that seemed fairly successful.by psb217 10y agoAlphaGo relied heavily on (supervised) pretraining, and that seemed fairly successful.