Mistakes happen. I have learned to be extraordinarily skeptical of extraordinarily good results (first and foremost, of my own results).
Every time I have gotten extraordinarily good results on a ML problem before, it turned out to be an issue with the experimental setup.
-
-
-
I'm super paranoid of my good results too. So much time spent double + triple checking, expecting a bug to appear each time! -_-
- Show replies
New conversation -
-
-
Ah yes, that unique mixture of excitement and dread when your results indicate a possible breakthrough...
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
exactly what I'm facing now. Model implemented in Keras though. makes debugging and verifying a whole lot easier
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
enjoy the moment... Then look for the bug!
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@chris_brockett that is such an annoyingly recurring theme that i started fearing really good results -
Probably a good plan. It also saves much embarrassment later if you don't announce them too soon. :)
End of conversation
New conversation -
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.