I am currently focusing on learning theory in a broader sense. I am particularly interested in how different machine learning problems are interconnected and characterized. Lately, reasoning about supervised learning and data corruption led to a project about Markovian corruption and its associated mitigations. Overall, I aim to find connections across fields, to allow better communication and formal understanding.
I am also keen on studying observed phenomena, e.g. those arising in neuroscience or social sciences, as well as thinking about issues of fairness and privacy. My first steps in academia moved in this direction by using network science to study social behavior, i.e. with influence-spreading models, collective action, and community/ranking structures.
Image Copyright © Nathan W. Pyle