Theoretical Grounding of Out-Of-Distribution Detection With Reinforcement Learning Optimizer

Salimeh Sekeh, Xin Zhang

Out-of-distribution (OOD) detection in dynamic open-world environments requires a model to continually adapt to evolving data distributions while generalizing to covariate-shifted…