Xiao-Bo Huang

Intelligent vulnerability assessment system for college students mental health based on few-sample reinforcement learning approach

It is suggested that combining few-sample meta-learning with metareinforcement-learning-based sequential decision-making can improve vulnerability assessment under scarce, evolving data relative to conventional static and non-meta-adapted approaches, with an accuracy-only reward recommended as the practical default.