The paper invetigates an interesting question, how to guarantee safety while being efficient. This is a very important problem. If we choose of be inefficient or conservative, the safety is very easy. For example, the autonomous vehicle...
I decide to share some random useful resources I collected during my research and personal life HERE. In future, I will also append my own reading notes for the research-related topics. Feel free to shoot me an email if you would like to comment on these resources! Read more about 09/2023: Sharing useful resources on misc
Two papers got accepted recently. Congratulations to all co-authors!
One paper is about multi-agent Bayesian optimization accepted to IROS 2023. In the multi-agent Bayesian optimization, a challenging question is that how to let the agents collaborate well on learning about the unknown objective function. We solve the batch queries by maximizing the joint information gain about the function maximum. We also conduct experiements with source seeking experiments using the TurtleBot3. Check out our...
Check our new preprint paper about multi-agent Bayesian optimization! Here is the link.
Multi-agent Bayesian optimization is an efficient approach for optimizing black-box functions which are difficult to evaluate. We use multiple agents to collaborate to find the optimum by maximizing the joint information gain. We conduct experiments on robotic experiments.
The paper I collaborated with Dongjie Yu (Master student at Tsinghua University), Reachability Constrained Reinforcement Learning, was accepted at ICML 2022. Congratulations to all coauthors!
