1 paper
Qingyang Zhang, Yatao Bian, Xinke Kong +2
Machine learning models must continuously self-adjust themselves for novel data distribution in the open world. As the predominant principle, entropy minimization (EM) has been pro…