Bernard J Koch

Hi Ywandung, Definitely not selection bias. I guess this would literally be bias from a statistical learning perspective (i.e., error in-sample). If your learning rate is too high, you risk...

Hi Ywandung, I really appreciate the feedback! I think I've fixed this issue, let me know! -B

Glad you enjoyed it. :) I haven't looked at these models particularly recently, so could you clarify for me which part is the bottleneck?

Hmm, those calculations are only for calculating nearest neighbors in representation space for validation. A couple quick solutions: 1) Skip them, they are only used in validation. Do it periodically...

Hi, Thank you for your interest! I had not planned to add additional tutorials at this time; so busy with other things. But hopefully this is helpful: 1. Should be...

Sounds good. Personally, I would've dropped the two works methods and just made one list of entities method, but wanted to make it backwards compatible with current methods. Also, no...

Well if it were me, I would probably remove all entity-specific methods and just keep get list of entities, not make it a private method, with an argument for entity...

Last couple of thoughts: -Institutions, Venues, and Author fields need to be expanded (but methods have to work with dehydrated values) -There are a couple supplementary objects (i.e., years, ids)...

Hi Dennis, If you give me less than an hour, I'm going to implement another branch. This one breaks current works functionality by generalizing for entities. The full set of...

Hi Dennis, I'm pretty much done with my rewrite in last major push. This does break existing get_works, but I think it does make API quite robust. I also changed...