Closed sparcycram closed 7 years ago
Is there anyway to change the activationfunction on dnn in caret?
I have tried using activationfun = "linear" from package deepnet in both tuneGrid = expand.grid(...., activationfun = "linear") and in trControl.
Neither worked. Am I doing something wrong or is this functionality not available?
Thanks
The method isn't parameterized for changing that right now. You can use the existing code to create a custom method. Pull requests are always welcome too =]
Is there anyway to change the activationfunction on dnn in caret?
I have tried using activationfun = "linear" from package deepnet in both tuneGrid = expand.grid(...., activationfun = "linear") and in trControl.
Neither worked. Am I doing something wrong or is this functionality not available?
Thanks