Runtime failure anticipation for frozen reinforcement learning driving policies: latent risk prediction and calibration-free alarm-rate control
Source code, configuration, result files, figure and table masters, and the frozen policy checkpoints for the article of the same title. Reproduction: python harness9/tables.py and python harness9/figures.py rebuild every table and figure in the article from the stored result files; the simulator is not re-run. This wo...
Miroslav B. Milovanovic, Jianxun Cui, Dragan Antić et al.
· Zenodo (CERN European Organi... · 0 citations