Quantum Machine Learning: A Framework for Near-Term Quantum Computing
Los Alamos National Laboratory
Machine learning has enabled task-oriented programming of classical computers, where the computer itself finds the optimal algorithm to accomplish a task that the user specifies. A natural question is: Could we do the same for quantum computers? After all, quantum algorithms are non-intuitive to the human mind, and generally it is difficult to discover new quantum algorithms. In this talk, I will review methods for task-oriented programming of gate-based quantum computers, a field that has been called quantum machine learning (QML). Examples of tasks for which researchers have proposed QML methods include finding quantum [UTF-8?] circuits that prepare a Hamiltonians ground state, factor numbers, diagonalize quantum states, perform error correction, and simulate quantum dynamics. The QML framework may enable the first practical applications of near-term quantum computers, and will likely impact many areas of physics research, ranging from condensed matter physics to quantum foundations.