Quantum circuit compilation by genetic algorithm for quantum approximate optimization algorithm applied to MaxCut problem

The Quantum Circuit Compilation Problem (QCCP) is challenging to the Artificial Intelligence community. It was already tackled with temporal planning, constraint programming, greedy heuristics and other techniques. In this paper, QCCP is formulated as a scheduling problem and solved by a genetic alg...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Swarm and evolutionary computation Ročník 69; s. 101030
Hlavní autoři: Arufe, Lis, González, Miguel A., Oddi, Angelo, Rasconi, Riccardo, Varela, Ramiro
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier B.V 01.03.2022
Témata:
Genetic algorithm
Makespan
Optimization
Quantum circuit compilation
Scheduling
Quantum circuit compilation
Scheduling
Makespan
Genetic algorithm
Optimization
ISSN:2210-6502
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The Quantum Circuit Compilation Problem (QCCP) is challenging to the Artificial Intelligence community. It was already tackled with temporal planning, constraint programming, greedy heuristics and other techniques. In this paper, QCCP is formulated as a scheduling problem and solved by a genetic algorithm. We focus on QCCP for Quantum Approximation Optimization Algorithms (QAOA) applied to the MaxCut problem and consider Noisy Intermediate Scale Quantum (NISQ) hardware architectures. Based on the fact that these algorithms apply a set of basic quantum operations repeatedly over a number of rounds, we propose a genetic algorithm approach, termed Decomposition Based Genetic Algorithm (DBGA), that in each round extends the partial solutions obtained for the previous ones. DBGA is compared to the state of the art across a set of conventional instances. The results of the experimental study provided interesting insight in the problem structure and showed that DBGA is quite competitive with the state of the art. In particular, DBGA outperformed the best current method on the largest instances and provided new best solutions to most of them.
ISSN:2210-6502
DOI:10.1016/j.swevo.2022.101030