Optimizing annealing parameters using genetic algorithms

ORAL

Abstract

Quantum annealing is an emerging technology that has the potential to be useful for solving combinatorial optimization problems such as maximum satisfiability. Prior work showed that annealing times [1] and inclusion of pauses [2] can significantly impact the probability of obtaining exact solutions. Less is known about leveraging other parameters such as flux biases implemented on the most recent D-Wave 2000Q annealer. We seek to bridge this gap by using genetic algorithms to select parameters. We explore various statistical measures, such as stochastic for ranking settings. Evaluations are performed using the Ames Research Center annealer.

[1] E. Crosson, arXiv:1401.7320
[2] J. Marshall, PhysRevApplied.11.044083

** We are grateful for support from NASA Ames Research Center, the Intelligence Advanced Research Projects Activity (IARPA), via IAA 145483, and the AFRL Information Directorate under grant F4HBKC4162G001. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of IARPA, or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purpose notwithstanding any copyright annotation thereon.

–

Presenters

  • Samuel Stromswold

    • QuAIL, NASA Ames Research Center

Authors

  • Samuel Stromswold

    • QuAIL, NASA Ames Research Center

  • Filip Wudarski

    • USRA - Univ Space Rsch Assoc

  • Eleanor Rieffel

    • Quantum AI Lab, NASA Ames Research Center
    • QuAIL, NASA Ames Research Center
    • NASA Ames Research Center, Quantum AI Lab (QuAIL)
    • NASA Ames Research Center
    • QuAIL, NASA