Statistical Limits of Supervised Quantum Learning
Abstract
Within the framework of statistical learning theory it is possible to bound the minimum number of samples required by a learner to reach a target accuracy. We show that if the bound on the accuracy is taken into account, quantum machine learning algorithms for supervised learning---for which statistical guarantees are available---cannot achieve polylogarithmic runtimes in the input dimension. We conclude that, when no further assumptions on the problem are made, quantum machine learning algorithms for supervised learning can have at most polynomial speedups over efficient classical algorithms, even in cases where quantum access to the data is naturally available.
Cite
@article{arxiv.2001.10477,
title = {Statistical Limits of Supervised Quantum Learning},
author = {Carlo Ciliberto and Andrea Rocchetto and Alessandro Rudi and Leonard Wossnig},
journal= {arXiv preprint arXiv:2001.10477},
year = {2020}
}
Comments
v3: 6 pages, journal version, title changed (previous title "The Statistical Limits of Supervised Quantum Learning"), other minor improvements; v2: 6 pages, title changed (previous title "Fast quantum learning with statistical guarantees"), format changed to two-columns, typos corrected, remarks that better clarify the limitations of our analysis added