Quantum optimization: from quantum annealing to reinforcement learning
Quantum optimization is the branch of quantum information that focuses on algorithms and techniques for ground-state search, based on quantum hardware and software. Indeed, low energy properties are not only interesting for condensed matter and quantum chemistry but are also useful to solve classical optimization problems. The general aim of quantum optimization is to exploit quantum mechanics to achieve the so-called "quantum supremacy" over analogous classical algorithms. In this talk, I will review quantum adiabatic computation, one of the basic tools of quantum optimization, and its relation with a recent proposal that is now at the center of intense research activities, the quantum approximate optimization algorithm (QAOA). Optimization algorithms are also strongly intertwined with optimal control techniques, which are essential to achieve good performance. A standard tool in classical control problems is reinforcement learning and I will discuss how it can be applied to QAOA.