US 12,468,977 B2
Uncertainty aware parameter provision for a variational quantum algorithm
Edward Oliver Pyzer-Knapp, Runcorn (GB); Mario Motta, San Jose, CA (US); and Michael Johnston, Dublin (IE)
Assigned to INTERNATIONAL BUSINESS MACHINES CORPORATION, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Jul. 16, 2021, as Appl. No. 17/378,437.
Prior Publication US 2023/0012699 A1, Jan. 19, 2023
Int. Cl. G06N 20/00 (2019.01); G06N 10/00 (2022.01)
CPC G06N 20/00 (2019.01) [G06N 10/00 (2019.01)] 19 Claims
OG exemplary drawing
 
1. A system, comprising:
a memory that stores computer executable components; and
a processor that executes the computer executable components stored in the memory, wherein the computer executable components comprise:
a decision component that determines, based upon an uncertainty prediction regarding usability of a defined parameter that has been output from a machine learning model, whether to employ the defined parameter for running a variational quantum algorithm;
a training component that trains the machine learning model by employing a central data store having data related to the variational quantum algorithm, wherein the training comprises training a machine learning model to be uncertainty aware by basing the machine learning model on a natively uncertainty aware machine learning algorithm; and
an Ansatz component comprises a set of quantum circuits with one or more free parameters, and approximates a quantum state of interest in which the one or more free parameters take optimal values, wherein the decision component outputs feedback to the Ansatz component to direct the Ansatz component to perform parameter optimization on the one or more free parameters using the defined parameter.