Hamza Keurti

Bio: Hamza Keurti is a PhD fellow at Max Planck ETH Center For Learning Systems. Prior his PhD, he obtained a Diplome d’Ingénieur (~MEng) at Centrale Paris with a double degree (MSc) in Advanced Computing at Tsinghua University. He also worked in software engineering at SLB. His research lies at the intersection of Machine Learning, Geometry and Compositionality. He is interested in learning  « good’’ representations of data which recover the compositional structure of the world. He draws inspiration from developmental psychology and places sensorimotor contingencies as a key element to learning such representations. In general, he is interested in avoiding the misspecification of models and representation structure in machine learning.