Advantage-Filtered Behavioral Cloning for Offline Continuous Control
This is the code for my 2021 undergrad thesis at the University of Virginia, "Advantage-Filtered Behavioral Cloning in High Noise Datasets" [Official Thesis Version], [Conference-Style (Shorter, ArXiv) Edit].
If you are interested in this code, please note that I have much more usable versions of this method implemented in my deep_control and super_sac repositories.