"KERAS model and Sliding Window Validation"

chinrenjiechinrenjie Member Posts: 8 Contributor I
edited June 2019 in Help

I am using sliding validation window and fix the training window width as 100 and testing window width as 100.
The total datasets that I have is 9000.
I want to develop a prediction model using the convolutional neural network in Keras Model.
I get the error as shown below. Can anyone please give some advice? Your help is highly appreciated.
Thank you.




  • Options
    rfuentealbarfuentealba Moderator, RapidMiner Certified Analyst, Member, University Professor Posts: 568 Unicorn

    Hi @chinrenjie


    On your right process (the testing part), just don't connect the mod port from the Apply Keras extension to the output of the process. This happens because mod is a model and what the res (or, in this case, ave) port is expecting is a data set or some other kind of object (in this case: averagable), and it cannot process a model.


    Whenever you have doubts, just put the mouse pointer on top of the semicircle. Something like this will show you what kind of data does it need (and if there are errors to be fixed):


    Screen Shot 2018-06-29 at 05.08.01.png

    Hope this helps.


  • Options
    chinrenjiechinrenjie Member Posts: 8 Contributor I


    I have tried to not connect the mod port to the output, the similar error happens. May I get advice regarding this issue? Thank you very much.

    Ren Jie




Sign In or Register to comment.