Abstract:We explore strategies aimed at reducing the amount of computation, both quantum and classical, required to run the Quantum Approximate Optimization Algorithm (QAOA). First, following Wurtz et al. [<a class="link-external link-http" href="http://Phys.Rev" rel="external noopener nofollow">this http URL</a> A 104:052419], we consider the standard QAOA with instance-independent "tree" parameters chosen in advance. These tree parameters are chosen to optimize the MaxCut expectation for large girth graphs. We provide extensive numerical evidence supporting the performance guarantee for tree parameters conjectured in [<a class="link-external link-http" href="http://Phys.Rev" rel="external noopener nofollow">this http URL</a> A 103:042612] and see that the approximation ratios obtained with tree parameters are typically well beyond the conjectured lower bounds, often comparable to performing a full optimization. This suggests that in practice, the QAOA can achieve near-optimal performance without the need for parameter optimization. Next, we modify the warm-start QAOA of Tate et al. [Quantum 7:1121]. The starting state for the QAOA is now an optimized product state associated with a solution of the Goemans-Williamson (GW) algorithm. Surprisingly, the tree parameters continue to perform well for the warm-start QAOA. We find that for random 3-regular graphs with hundreds of vertices, the expected cut obtained by the warm-start QAOA at depth $p \gtrsim 3$ is comparable to that of the standard GW algorithm. Our numerics on random instances do not provide general performance guarantees but do provide substantial evidence that there exists a regime of instance sizes in which the QAOA finds good solutions at low depth without the need for parameter optimization. For each instance studied, we classically compute the expected size of the QAOA distribution of cuts; producing the actual cuts requires running on a quantum computer.

Qudit inspired optimization for graph coloring

Exploring the Potential of Qutrits for Quantum Optimization of Graph Coloring

Quantum approximate optimization algorithm for qudit systems

Encoding trade-offs and design toolkits in quantum algorithms for discrete optimization: coloring, routing, scheduling, and other problems

Quantum Graph Optimization Algorithm

Quantum Circuit Optimization through Iteratively Pre-Conditioned Gradient Descent

Iterative quantum algorithm for combinatorial optimization based on quantum gradient descent

Strategies for running the QAOA at hundreds of qubits

Quantum-Informed Recursive Optimization Algorithms

Experimental implementation of quantum greedy optimization on quantum computer

A benchmarking study of quantum algorithms for combinatorial optimization

Optimization Applications as Quantum Performance Benchmarks

A Quantum Approximate Optimization Algorithm for continuous problems

Mapping graph coloring to quantum annealing

Mixed-Dimensional Qudit State Preparation Using Edge-Weighted Decision Diagrams

Optimizing a Polynomial Function on a Quantum Processor

Energy Landscapes for the Quantum Approximate Optimisation Algorithm

Recursive Quantum Relaxation for Combinatorial Optimization Problems

Graph decomposition techniques for solving combinatorial optimization problems with variational quantum algorithms

Q-CHOP: Quantum constrained Hamiltonian optimization

QAOA of the Highest Order