4 ms·
This conclusion feels like saying more CPU and memory are better. Seems obvious that more moves allows matching to have more nuance, but I guess cool that someo
by AmericanBlarney 5y ago
This conclusion feels like saying more CPU and memory are better. Seems obvious that more moves allows matching to have more nuance, but I guess cool that someone proved it.
- nazgul17 5y agoFrom what I understand, it says that more parameters are good. This wasn't obvious before this paper: you can fit a polynomial instead of a neutral net, but adding parameters wouldn't help with robustness in that case: the polynomial would become more and more jagged.
- ska 5y ago> Seems obvious that more moves allows matching to have more nuance, This really has to be balance against overfitting. The key problem in ML is generalization, and lots of things improve training performance while making that worse.