Exploring quantum-informed recursive optimization algorithms on Amazon Braket

This article discusses quantum-informed recursive optimization (QIRO) algorithms for solving constrained combinatorial optimization problems, such as the maximum independent set (MIS) problem, using a hybrid quantum-classical approach. The key points are:

Specifically, the article covers:

• Challenges of encoding constrained optimization problems in the standard QUBO (quadratic unconstrained binary optimization) formulation for quantum optimization.
• The QIRO algorithm, which recursively simplifies the problem using correlations obtained from quantum resources like QAOA or quantum annealing.
• Experimental results showing QIRO informed by neutral atom quantum hardware (QuEra Aquila) outperforming classical benchmarks for the MIS problem.
• Potential improvements to QIRO, such as tailoring update rules to specific problems and improving the quantum component.