Each year, a select group recognized by Google researchers and their institutions as some of the most promising young academics in the world become recipients of Google PhD Fellowships. These Fellowships directly support graduate students as they pursue their PhD, as well as connect them to a Google Research Mentor. Amir-Hossein Karimi, of the Max Planck Institute for Intelligent Systems and CLS, is a 2021 recipient of a Google PhD Fellowship in Machine Learning.
The research interests of Amir-Hossein Karimi focus on ethical and collaborative AI, particularly using AI systems to safely augment human intelligence and foster collaboration. As predictive models are increasingly being used to support decision-making in consequential social contexts (e.g., loan approval, hiring, and healthcare), there is mounting social and legal pressure to provide explanations that help the affected data-subjects not only to understand why a prediction was output, but also how to act to obtain a desired outcome from an automated system. During his PhD, Amir aims to study, design, and deploy methods to address these questions, with the aim of offering algorithmic recourse to affected individuals. To date, Amir’s PhD research has led to publications at AISTATS2020 and FAccT 2021, as well as a NeurIPS 2020 Spotlight.
Amir-Hossein Karimi joined CLS as a doctoral fellow in 2018 and is supervised by Prof. Bernhard Schölkopf, Director of Department of Empirical Inference at the Max Planck Institute for Intelligent Systems Tübingen, Germany and Affiliate Professor, Department of Computer Science at ETH Zurich, Switzerland, together with Prof. Dr. Isabel Valera, formerly of MPI-IS and currently Professor on Machine Learning at the Department of Computer Science of Saarland University. From September 2020 – August 2021, Amir carried out a research exchange in the Data Analytics group of Prof. Dr. Thomas Hofmann at the Department of Computer Science at ETH Zurich. Amir is a doctoral candidate at ETH Zurich and aims to defend his doctoral degree by the end of 2022.