Investigate the influence of different warm-start and mixing operators on QAOA forweighted-SAT problems
Context: The Quantum Approximate Optimization Algorithm (QAOA) is used as a versatile hybridalgorithm in the current NISQ-era of quantum computing. We set up a demonstrator JupyterNotebook to show how to solve boolean 2-satisfiability (2-SAT) and more complex weighted-2-SAT instances. These NP-hard problems need to be solved regularly, for example in the automotive industry.
Goal: Using our demonstrator as a baseline, replace the Initialisation and Mixer parts of the QAOAalgorithm with different implementations. A first candidate would be a Grover’s search inspired warmstart, followed by a Grover diffusion operator. More variations may be researched and implemented. Evaluate your implementation with commonly used quantum metrics (circuit width,depth), as well as determining the quality of the algorithm’s results.
Requirements: Linear algebra, Boolean Logic, Python, (Basic knowledge of gate based quantumcomputing is beneficial but can be acquired during the thesis)