https://www.youtube.com/watch?v=_Fmkqwrt_Gk&t=489s --> ‘Preprocess image data’ workflow
I found the workflow at KNIME Hub -> knime -> Spaces -> Examples -> 04_Analytics -> 14_Deep_Learning -> 02_Keras -> 04_Cats_and_Dogs but the Python node is missing.
I am trying to do something similar to what it’s done in the video, so it would be nice to see this code as the tutorial was very informative.
Can anyone tell me where I can find it or does anyone have it?
Thanks a lot
Hi @leo4,
David slightly modified the workflow from the example. You can find the workflow used in the workshop at his HUB space: https://hub.knime.com/davek/spaces/Public/latest/DL_Workshop_ImgAna_2020/CatsDogsClassification/.
Thank you very much, Benjamin
Sorry to bother @bwilhelm! Just one question: how can you see if you are overfitting? I see no possibility of generating a test-train loss plot. Do you think it makes sense to generate a validation set?
I can only see the loss and accuracy plots for the training data alone. It’s already a good indicator because I see that a plateau is reached, so it doesn’t make sense to do all the epochs I did. I will try to check the earlystopping box. I had only seen this with validation loss, but I guess it makes sense this way, too
This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.