search

deepchem / deepchem

Democratizing Deep-Learning for Drug Discovery, Quantum Chemistry, Materials Science and Biology
https://deepchem.io/
MIT License
5.39k stars 1.66k forks source link

Attention Recurrent Comparators #809

Open rbharath opened 7 years ago

rbharath commented 7 years ago

https://medium.com/@sanyamagarwal/understanding-attentive-recurrent-comparators-ea1b741da5c3

We might want want to add an implentation to dc.metalearning

lilleswing commented 7 years ago

While the results are cool -- I am failing to see how they transfer to Chemistry problems.

The problem that these networks solve are One Of N classification where new classes are added after training time.

lilleswing commented 7 years ago

Also I did a cursory pass at attempting to apply this to a generalized multi-task classification/regression problem.

rbharath commented 7 years ago

The original matching networks paper was also for one of N w/ new classes added in :-)

I haven't yet taken a close look, but I suspect the attention recurrent comparators can be ported over similarly to how the matching networks and iterative recurrent comparators were. Will try to take a closer look when I get a chance.

rbharath commented 6 years ago

I'm marking this a good intermediate contribution for those with deep learning expertise. Also, I suspect this implementation might be easier done with TensorFlow Eager than with graph structured tensorflow (but could be wrong on this hunch).