Overview

Directory structure

root

• code
  • backpropagation_code
    • general_purpose_notebook.ipynb the main notebook for backprop networks to train models, evaluate, create figures, etc
    • model_comparisons contains json files comparing the scores obtained from different inputs
    • benchmark_code code for running benchmarks as well as the results obtained from them
  • modules Python modules with finished functions used by other notebooks/scripts
  • non_network_notebooks Notebooks used to generate non-neural network results, e.g. modify galaxy catalogues
  • pso_code
    • pso_general_purpose_notebook.ipynb The main notebook for backprop networks to train models, evaluate, create figures, etc
  • scripts
    • reinforcement_training.py Script used for initiating the reinforcement training