Speaker: Finley Quinton, NTNU

Title: Expressive and Constraint-Preserving QAOA for the Minimum Vertex Cover with Optimized Initializations

Abstract: Our work investigates advanced Quantum Approximate Optimization Algorithm (QAOA) strategies for the Minimum Vertex Cover (MVC) problem on weighted graphs, emphasizing constraint-preserving mixers and their performance against traditional penalty-based approaches. We analyze both the standard constraint-preserving mixer and a multi-angle variant, exploring how careful parameter initialization and constraint enforcement can improve convergence and solution quality. In the multi-angle setting, we propose strategies to reduce the number of parameters to optimize, aiming to improve scalability and mitigate barren plateaus. Extensive experiments on different weighted graph instances highlight the advantages of these approaches, demonstrating more reliable performance and better alignment with problem-specific structures.