New tutorial on Recurrent Neural Networks (RNN)

[kxk302]

This is a new tutorial for Statistics and Machine Learning topic. I have uploaded the data to Zenodo, and have included the workflow as well. I followed the instructions/tutorials on GTN to write this tutorial. Please review and let me know if any corrections are needed. Thank you!

P.S. As suggested, I used yEd to create the figures in this tutorial. Was wondering if its ok to add the .graphml files in the images folder (or a new folder), in case the figures need to be updated later?

[hexylena]

@anuprulez as one of our newest maintainers would you have a chance to look at this? Is the content ok, is everything explained well enough for beginners, etc?

[anuprulez]

@hexylena Yes, I will have a look at this tutorial a bit later today. Thanks!

[anuprulez]

@kxk302 thanks for the tutorial. I have added some comments and will add some more.

[kxk302]

@kxk302 thanks for the tutorial. I have added some comments and will add some more.

Thanks @anuprulez. I revised the tutorial based on your comment and pushed it to my branch.

[anuprulez]

@kxk302 Thanks for updating the tutorial. I tried running the tutorial locally but I ran into some problems. Can you address it? This is my history: https://usegalaxy.eu/u/kumara/h/rnn-tutorial

Edit: I imported the workflow for the tutorial. Please add labels for each dataset which will make it easier to know in which input field which dataset to add. In the following example, I am not sure where to attach X_train or y_test.

Thanks for working on it :)

[kxk302]

https://usegalaxy.eu/u/kumara/h/rnn-tutorial

Added name/label to all inputs. Looking at workflow issue next.

[kxk302]

@anuprulez I imported the latest workflow and it completed successfully. The only thing I had to do was convert y_test and y_train to tabular (Updated the instructions so its clear).

[kxk302]

Thank you for all the helpful feedback!

[shiltemann]

@kxk302 this is very cool, thanks! With respect to your question, yes, please add the .graphml source file in case we want to update the image later :+1:

[kxk302]

@kxk302 this is very cool, thanks! With respect to your question, yes, please add the .graphml source file in case we want to update the image later 👍

Will add .graphml files. Thanks @shiltemann!

[hexylena]

Build failure is due to #2333

[kxk302]

@anuprulez and @hexylena, Was wondering if there is any other change that you suggest to the tutorial? (if not, can this be merged?) Thanks!

[anuprulez]

I will run it in my local this week and provide comments if any. Thanks @kxk302

[kxk302]

I will run it in my local this week and provide comments if any. Thanks @kxk302

Thanks @anuprulez!

[kxk302]

@hexylena I pushed my changes for your comments.

[anuprulez]

@kxk302 thanks for making the changes. I checked the tutorial in my local and all the steps are working fine. However, I have one minor suggestion - I do not see the final confusion matrix plot in the tutorial. Do you think it would be good to add this plot and write one or two sentences about predictions/plot?

@hexylena Apart from this, it looks good to me :+1:

[kxk302]

@kxk302 thanks for making the changes. I checked the tutorial in my local and all the steps are working fine. However, I have one minor suggestion - I do not see the final confusion matrix plot in the tutorial. Do you think it would be good to add this plot and write one or two sentences about predictions/plot?

@hexylena Apart from this, it looks good to me 👍

@anuprulez I think its a good idea to add the confusion matrix with some explanation. Will add it and update this ticket when its pushed.

[kxk302]

@anuprulez I just pushed my change that adds a Confusion Matrix image and analysis.

[anuprulez]

@kxk302 @hexylena :+1:

[hexylena]

Thanks so much @kxk302! It'll be online in ~5-10 minutes.

[kxk302]

@anuprulez @hexylena Thank you both very much for all the constructive feedback! Anup, I have a couple more ML tutorials in the pipeline so we will meet again soon :-)

[bgruening]

That is pretty cool! Thanks @kxk302 and @anuprulez!

[anuprulez]

@kxk302 thanks :)