Workshop on: Machine Learning, Quantum Acceleration and Robust Quantum Systems
November 18-19, 2019, Luskin Conference Center
Quantum algorithms offer the prospect of unprecedented advantages in solving problems in statistical mechanics, condensed matter and materials physics, quantum chemistry and cryptography. The current generation of quantum computers based on superconducting qubits or trapped ions is limited in scope due to the small number of qubits that can be entangled with high fidelity. Such scalability problems prevent the implementation of practically useful quantum algorithms, most of which require much larger numbers of qubits. We are thus constrained, at this time, to work with unscalable architectures, leaving out many applications and limiting the scope of the research. Nonetheless, current problems being addressed are in the realm of statistical physics of many-body problems and phase transitions, quantum-assisted machine learning (ML), quantum-assisted sensing and the design of robust control algorithms for quantum systems. In recent years, ML has gained popularity with the development of deep learning applications, which are capable of simulating the behavior of complex systems by learning to synthesize knowledge from large amounts of training data. This has led to the search for methods that exploit the power of ML to enhance the performance of quantum systems and conversely, the application of quantum systems to enhance the performance of ML algorithms. Moreover, recent developments in explainable AI may lead to the rational design of smaller, quantum-assisted algorithms with applications to simpler and more tractable problems.
The workshop specifically addressed:
- Quantum algorithms for accelerating ML
- ML applied to the design, characterization and control of quantum systems
- Nonstandard (emerging) statistical learning algorithms
- Quantum Control and Landscape Theory
- Parallel and distributed algorithms for quantum processors
- Experimental quantum systems for QML
- Applications to statistical mechanics, condensed-matter physics, robust quantum control, control landscape theory or quantum-assisted sensing
- Deriving quantum advantages for ML problems
- Using data driven methods for characterization and control of quantum architecture platforms
With ML reaching its limits due to the limitations in computational power, the combination of ML with quantum algorithms is timely, and offers new possibilities for breaking existing barriers to computation. This workshop brought together experts from math, physics, engineering and chemistry, in order to foster interactions as well as encourage collaborations between academia and industry.
Confirmed speakers: Herschel Rabitz (Princeton), Jarrod McClean (Google Venice), Hari Krovi (Raytheon), Igor Markov (Univ of Michigan), Jonathon Dubois (LLNL), Gian Guerreschi (Intel), Dmitri Maslov (IBM Q), Dave Wecker (Microsoft), Dan Lidar (USC), Seth Lloyd (MIT), Roger Melko (Waterloo), Maria Schuld (Xanadu Toronto), Michael D Schneider (Lawrence Livermore National Laboratory).
Organizers: Vwani Roychowdhury (UCLA), Louis Bouchard (UCLA), Eric Hudson (UCLA), Kang Wang (UCLA), Mark Gyure (UCLA), Herschel Rabitz (Princeton), Jarrod McClean (Google Venice), Gian Guerreschi (Intel).