@abe_asfaw Probably the best way to pose this question is "Was the quantum computing field sufficiently subsidized to support the huge flock of people coming in?" Or "Was X a safe and substantial place for them to express themselves?" etc.
I just can't believe I am reading review papers and they still talk about unresolved bandgap values in well-known materials. What? Is accurate band-gaps not sexy enough to be funded?
Unveiling Intrinsic Bulk Photovoltaic Effect in Atomically Thin ReS2 | Nano Letters https://t.co/tEdR6NjNzg
Our article on shift conductivity in ReS2 is out in ACS nanoletters. I feel so lucky to be part of this wonderful collaboration.
So, all important contributions in applications of #MachineLearning basically need to be a balance between efficient use of resources and usefulness of predictions. This is the first barrier that needs to be overcome.
Another option is to find low-cost simulations to run thousands of times, and then train your NN on that. But, this usually reduces the significance of your NN predictions.
The way to get data is either through experiments or through simulations. Simulations also need computer power while experiments... well a lab and all expenses covered.
#data#machinelearning
By now, many people know the basics of Neural Networks (IT people even long before, since it is a course in Universities). What is less obvious to the beginner is the two conditions needed: Computer power and data. Data is the currency of today.
https://t.co/7okstTbhxv