Embedded Files
Challenges in realizing AI for social impact
Challenges in realizing AI for social impact
Several challenges arise because it is AI for social impact. For example, we need to plan on collaborating with humans from data collection to deployment. We may also have limited, noisy data, which amplifies uncertainty.
Our lab seeks to develop novel techniques in the fields of multi-agent systems and machine learning to address these challenges and help foster safe deployment of AI systems in society.
Human-AI Collaboration
Human-AI Collaboration
Humans are a vital part of deployed systems, and may play different roles in them. How can humans and AI systems best collaborate?
Past Projects:
Uncertainty
Uncertainty
In using real-world data, there is often noise and limited data. How can we model uncertainty to capture real-world complexities?
Past Projects:
Page updated
Google Sites
Report abuse