Closed matteopariset closed 5 years ago
Hi @matteopariset,
Sorry for the confusion -- those scripts largely deal with our internal workflow and require access to the full UK Biobank dataset. As such they're not really meant for public use.
The weak supervision part of the paper is covered in these two Jupyter notebooks . We've also provided a demo of our end model (CNN-LSTM) that uses labeled MRI data here.
We're in the process of putting together a new end-to-end demo that should be easier to use and is built using our upcoming Snorkel v0.9 release. That should come online by the end of August.
Thanks for the interest!
Thank you for the very fast reply. I look forward to trying out the new material
I'm trying to run the
WeakSupervised.sh
script but I don't fully understand what data should be used (for example in SEMI parameter inscaleup_base.sh
) and where I can find the fileANON_unknown_hack3_gender.csv
Thanks in advance