These are three Python notebooks from my work with the Magnino team in summer 2020

quiver_autoencoder:
a. Using the 2i2o dataset provided by Kyle, train autoencoder NN
b. Test multiple different permutations of autoencoders to achieve high
model accuracy

2. fine_grain_dataset:
a. Load NN trained in quiver_autoencoder
b. Generate fine-grain dataset with random spin vectors
c. Pass new fine-grain dataset into NN, which outputs with dataset with
interpolated spin vectors
d. Visualize results

3. time_autoencoder:
a. Worked on towards end of summer—less definitive results
b. Given time dataset from Kyle, train autoencoder NN
c. Pass 5vect dataset into NN and investigate resultant predicted dataset