Re:All-optical machine learning using diffractive deep neura...
by h4b4n3r0 on Monday August 06, 2018 @11:28AM
(#17698443)
Nonlinearity is very important and is the only reason why neural nets can approximate arbitrary functions. You can’t do that with linear transformations alone. Though from briefly skimming the paper they do seem to achieve similar effects through phase modulation. Otherwise even MNIST would be out of the question.
[ Parent
| on HN ]
There are no comments on this comment yet.