Qiyu Wan
Papers
1
Total Citations
7
H-Index
1
About
Qiyu Wan is a researcher advancing the frontiers of probabilistic machine learning, with a primary focus on Bayesian Neural Networks (BNNs) and their deployment in safety-critical AI systems. Their most cited work, "Shift-BNN: Highly-Efficient Probabilistic Bayesian Neural Network Training via Memory-Friendly Pattern Retrieving" (2021, 7 citations), addresses a fundamental challenge in BNNs: the computational and memory overhead of training models that can quantify uncertainty. Wan’s key contribution lies in developing a memory-efficient training framework that retrieves patterns to accelerate probabilistic inference, making BNNs more practical for real-time applications like autonomous driving, medical diagnosis, and robotics. This work has been recognized for bridging the gap between theoretical uncertainty estimation and scalable deployment, earning citations from researchers in reliable AI and embedded systems. Wan’s research is particularly impactful for students and engineers seeking to integrate robust decision-making into resource-constrained environments, where traditional BNN training is often infeasible. By enabling efficient uncertainty-aware learning, Wan’s work supports the next generation of trustworthy AI, ensuring that models can not only predict but also know when they might be wrong—a critical capability for life-critical technologies.
Research Focus
Key Achievements
Top Papers
- 1