This entree is a part of our Meet the Fellow blog series, which introduces and highlights Faculty Fellows who have recently joined CDS
Meet CDS Assistant Professor/Faculty Fellow Umang Bhatt, who will join CDS this fall. A PhD candidate in the Machine Learning Group at the University of Cambridge advised by Adrian Weller, Umang will continue to pursue research in trustworthy machine learning, responsible artificial intelligence, and human-machine collaboration at NYU.
“Home to both a thriving tech ecosystem and pioneering efforts on regulating algorithm decision-making systems, New York City provides a vibrant research environment and a plethora of interdisciplinary collaborators,” said Umang. “For these reasons, I am excited to start my academic journey at NYU. I look forward to collaborating with CDS faculty and students to build and deploy machine learning systems that augment and complement human decision-makers!”
Umang’s PhD is funded by the Leverhulme Center for the Future of Intelligence (Trust and Transparency Initiative) via generous donations from DeepMind and the Leverhulme Trust. Motivated by applications in healthcare and criminal justice, Umang studies how to create algorithmic decision-making systems endowed with the ability to explain their behavior and adapt to a stakeholder’s expertise to improve human-machine team performance.
He develops methods grounded in information theory and probabilistic machine learning, while drawing from advances in cognitive science and psychology. “My research style includes convening stakeholders to understand gaps in the ecosystem, devising principled methods to address stakeholder needs, and running large-scale user studies to study the efficacy of proposed methods,” said Umang.
Some examples of Umang’s research include: “Explainable Machine Learning in Deployment” (ACM Conference on Fairness, Accountability, and Transparency 2020), “How Transparency Modulates Trust in Artificial Intelligence” (Patterns 2022), and “Eliciting and Learning with Soft Labels from Every Annotator” (AAAI Conference on Human Computation and Crowdsourcing 2023). His work has been covered in press (e.g., IEEE Spectrum, Amazon Science) and referenced in policy briefs (e.g., UK Parliament POSTnote, NIST).
In addition to working on his advanced degree, Umang is a Research Associate on the Safe and Ethical AI Team at the Alan Turing Institute. He is an Advisor at the Responsible AI Institute and has served in mentoring roles as a Thesis Co-Supervisor and Teaching Assistant at the University of Cambridge.
In 2022, he was awarded a J.P. Morgan AI PhD Fellowship and joined Harvard University’s Center for Research on Computation and Society as a Research Fellow. He has previously held fellowship positions with the Mozilla Foundation and the Partnership on AI.
Umang earned a joint bachelors-masters in Electrical and Computer Engineering at Carnegie Mellon University, where he was advised by José Moura and collaborated with Pradeep Ravikumar on explainable AI and Zico Kolter on automated pothole detection.
To view all our current faculty fellows, please visit the CDS Faculty Fellows page on our website.
By Meryl Phair