Michael D. Schneider, Group Leader for Astronomy & Astrophysics Analytics, Physics Division, Lawrence Livermore National Laboratory
Quantum Machine Learning using Gaussian Processes
Quantum computers may be transformative for a variety of computational tasks. We are pursuing machine learning methods for near-term quantum devices that may show advantage over classical computers. While classical unsimulatability of a quantum system is a necessary condition for quantum advantage, it is not sufficient as not all such quantum systems will be equally effective in prescribed computations. I will describe practical implementation of Gaussian Process based machine learning on existing quantum computers that can perform machine learning tasks at least as well as, and many times better than, the classical inspiration. By identifying our Gaussian Process problem with related methods in deep learning and signal processing, I will describe possible paths towards future advances in quantum machine learning.