Making assumptions about the structure of correlations in the data will *reduce* the expressiveness of the model (that's the point: dramatically reduce the search space to make search more data-efficient and faster). Which is why it is important to make *correct* assumptions.https://twitter.com/AmirRosenfeld/status/1033850972299321344 …
-
-
Good observation.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
A good approach but what do you think is the alternative? We almost use conv layers in all cases to promote the most discriminative features ..
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.