Differential Evolution as QAOA Optimizer for the Community Detection Problem
Clara Pizzuti · 2024
A hybrid quantum optimization algorithm that uses differential evolution as a gradient-free optimizer to solve the community detection problem is proposed. A problem-dependent quantum circuit is generated for a network and the search for the optimal parameters of the circuit is performed by evolving a population of individuals representing the quantum parameters. The approach is evaluated on networks synthetically generated whose division into two communities is known. The experimentation shows that the method can find the ground-truth communities for almost all the considered graphs, misplacing at most a few nodes for those networks whose community structure is not very clear.