Feedback from Phil

[PhilChodrow]

This looks very interesting! I really like that you were able to identify some candidate public data sets, as this was something I was initially worried about. Molecule vectorization looks like a very interesting task that I have never heard of! We will indeed be covering TensorFlow in lectures. So, one approach for you would be to focus on data acquisition and vectorization in the first few weeks, and then start thinking about deploying models once we cover TensorFlow in a week or two.

Overall, excellent! This is a very thoughtful proposal, and I look forward to seeing what you are able to produce.