Abstract: Can we exploit the tools of machine learning for tackling the complexity of building better quantum computers? In this talk, I will describe how one can use machine-learning techniques, specifically reinforcement learning, to address challenges in this domain. One very important area is quantum error correction, which is needed to fight against the unavoidable noise that would otherwise quickly render quantum computers useless. The complex feedback strategies needed in this domain can be discovered using model-free or model-based reinforcement learning. Besides discussing our theoretical advances in this area, I will also describe our recent collaboration with experimentalists in learning real-time feedback strategies for superconducting qubits. Co-sponsored by: Montréal Quantum Photonics Seminar Series Speaker(s): Florian Marquardt 6666 Rue Saint-Urbain, Montréal, Quebec, Canada