Anthony Gitter

There is a lot to like in that Mayr et al. 2018 paper. They were very thorough in exploring different models, hyperparameters, and chemical featurizations. However, they only consider ROC...

@evancofer my personal preference would be to create a new issue for each review we want to cite. That way we can briefly note why we want to include that...

@evancofer that would be cool! Want to open a new issue to discuss what that would look like?

Along with helpful tips, this rule could include how to assess whether deep learning is providing any benefits on your task. Or perhaps benchmarking is important enough to be a...

> What label should an issue like this have? "meta" or something else? I agree with "meta". It is proposing terminology to use consistently as opposed to a new rule...

I'm okay using training/tuning/testing if we also prepare readers for the terms they will encounter in machine learning or biomedical literature. There is precedent for training/tuning/testing in machine learning. The...

> More often than not, if a human can't see "it", a machine won't either What type of data are you thinking of here? I agree that we should counter...

I find the transcriptomics applications to be more convincing, even if Casey [has shown](https://doi.org/10.1101/385534) their dependence on parameters. In biochemistry / drug discovery, [this autoencoder](http://doi.org/10.1021/acscentsci.7b00572) is another example of finding...

I'm not sure how to articulate this, but dataset size is relative to the complexity of the problem. We've had problems where the input-output mapping is straightforward enough to learn...

Yes, that's similar to what I had in mind. Using more flexible labeling makes sense. Are these "citations" or some more general type of cross-reference? I haven't looked at what...