References & Citations
Quantum Physics
Title: An Adaptive Optimizer for Measurement-Frugal Variational Algorithms
(Submitted on 19 Sep 2019 (v1), last revised 6 May 2020 (this version, v3))
Abstract: Variational hybrid quantum-classical algorithms (VHQCAs) have the potential to be useful in the era of near-term quantum computing. However, recently there has been concern regarding the number of measurements needed for convergence of VHQCAs. Here, we address this concern by investigating the classical optimizer in VHQCAs. We introduce a novel optimizer called individual Coupled Adaptive Number of Shots (iCANS). This adaptive optimizer frugally selects the number of measurements (i.e., number of shots) both for a given iteration and for a given partial derivative in a stochastic gradient descent. We numerically simulate the performance of iCANS for the variational quantum eigensolver and for variational quantum compiling, with and without noise. In all cases, and especially in the noisy case, iCANS tends to out-perform state-of-the-art optimizers for VHQCAs. We therefore believe this adaptive optimizer will be useful for realistic VHQCA implementations, where the number of measurements is limited.
Submission history
From: Andrew Arrasmith [view email][v1] Thu, 19 Sep 2019 16:31:53 GMT (333kb,D)
[v2] Tue, 1 Oct 2019 17:24:04 GMT (388kb,D)
[v3] Wed, 6 May 2020 05:32:59 GMT (787kb,D)
Link back to: arXiv, form interface, contact.