Jiayu Yao
Papers
1
Total Citations
10
H-Index
1
About
Jiayu Yao is a researcher whose work sits at the intersection of Bayesian machine learning and trustworthy AI. Her most cited contribution, "Output-Constrained Bayesian Neural Networks" (2019), tackles a fundamental limitation of standard Bayesian neural networks: the difficulty of encoding prior knowledge about a model's behavior in function space rather than parameter space. Yao introduced a novel formulation that allows practitioners to impose functional constraints on model outputs in specific input regions, significantly enhancing the interpretability and safety of predictions. This work has garnered 10 citations and represents a key step toward building more reliable AI systems. Beyond this, Yao's research explores uncertainty quantification and robust decision-making under distribution shift, with applications in healthcare and autonomous systems. Her contributions are particularly valuable for researchers seeking to bridge the gap between Bayesian theory and practical, safety-critical deployments.
Research Focus
Key Achievements
Top Papers
- 1Output-Constrained Bayesian Neural Networks10 citations · 2019