This code trains a RDF classifier on good/bad behavior of an autonomous vehicles using scenarios derived from the California DMV disengagement reports.
agents/: Contains the code which controls the behavior of autonomous vehicle (ego) and adversaries.
cross_entropy/: Contains code to define various distributions as well as the cross entropy class which defines the primary cross-entropy loop.
scenarios/: Defines scenarios that inherit from the ScenarioClass base class.
main.py: The main program that generates the 0's/1's
process_labeled.ipynb: Jupyter notebook that processes the data generated and trains/evaluates the RDF classifier