
Daniel Flam-Shepherd
@FlamShepherd
Followers
107
Following
54
Media
3
Statuses
8
PhD student @UofTCompSci @VectorInst Working on machine learning for molecular design.
Toronto
Joined January 2019
We built a framework for fragment-based 3D molecular design using reinforcement learning. Scales to more than 100 atoms. Preprint: w/ @alex_zhigalin, @A_Aspuru_Guzik . Watch the agent design a molecule by placing fragments from Imatinib:
2
13
84
RT @HochreiterSepp: ArXiv LSTMs with SMILES and SELFIES perform better at generating complex molecules than graph….
0
31
0
RT @MarioKrenn6240: Deep Learning can help solving many questions in Science. But how does it think about the world?. We analyze deep gener….
0
9
0
RT @A_Aspuru_Guzik: At the interface of #machinelearning and #quantumoptics ! #generativemodels learn implicitly about entanglement and bui….
0
2
0
RT @NMcgreivy: Stellarators are fusion reactors which use an optimized twisting magnetic cage to confine 100 million ℃ plasma. Here, I use….
0
58
0