Quantum Inf Process (2015) 14:3933–3947
Protocol for secure quantum machine learning at a
· Seung-Woo Lee
Received: 19 April 2015 / Accepted: 27 July 2015 / Published online: 7 August 2015
© Springer Science+Business Media New York 2015
Abstract The application of machine learning to quantum information processing
has recently attracted keen interest, particularly for the optimization of control para-
meters in quantum tasks without any pre-programmed knowledge. By adapting the
machine learning technique, we present a novel protocol in which an arbitrarily ini-
tialized device at a learner’s location is taught by a provider located at a distant place.
The protocol is designed such that any external learner who attempts to participate in
or disrupt the learning process can be prohibited or noticed. We numerically demon-
strate that our protocol works faithfully for single-qubit operation devices. A trade-off
between the inaccuracy and the learning time is also analyzed.
Keywords Quantum computation · Quantum machine learning · Secure machine
Advances in quantum information science herald a new era of information technology.
Quantum information science has recently penetrated interdisciplinary science and
engineering ﬁelds. In particular, a current research topic is to adapt the basic idea
of machine learning for quantum information processing. Although “learning” is a
behavior of humans and other living things, a device or a machine can also learn a task
according to the theory of machine learning, which was developed as a subﬁeld of
artiﬁcial intelligence . In fact, the optimization of control parameters without any
Department of Physics and Astronomy, Center for Macroscopic Quantum Control, Seoul National
University, Seoul 151-747, Korea