>>12219183Just go into google scholar and search "machine learning materials design" or "machine learning molecules" or "machine learning proteins".
For drug design, we have a goal to find a molecule that best binds to a target protein. There are infinite possibilities, so we use genetic algorithms to artificially evolve molecules.
For materials physics, we use neural networks to approximate the energy of interaction between atoms so that we can predict material properties. This involves making a neural network that inputs atomic positions and spits out energy, so it's a regression problem.
Just some examples, but there are many more.
