 |
Center for Nonlinear Studies |
Machine learning applications are limited by computational power. Quantum annealing (QA) is a new computing paradigm that shows promise for solving complex optimization problems of the quadratic unconstrained binary (QUBO) format. In this talk, we discuss the application of QA to machine learning (ML) through experiments in natural language processing, seizure prediction, and linear separability testing. These experiments were performed on QA simulators and early-stage commercial QA hardware and compared to an unprecedented number of traditional ML techniques. We extend QBoost, an early implementation of a binary classifier that utilizes a quantum annealer, via resampling and ensembling of predicted probabilities to produce a more robust class estimator. To determine the strengths and weaknesses of this approach, resampled QBoost (RQBoost) is tested across several datasets and compared to QBoost and traditional ML. We find that early stage QA-enabled machine learning outperforms some, but not all traditional machine learning techniques. We consider these results encouraging given the very small number of quantum bits these early devices can make use of. Through these experiments, we provide unique insights into the state of quantum ML via boosting and the use of quantum annealing hardware that are valuable to institutions interested in applying QA to problems in ML and beyond.
Host: Nikolai Sinitsyn