Mendl, ChristianLorenz, Jeanette MiriamDragan, Theodora-AugustinaTheodora-AugustinaDragan2022-11-072022-11-072022https://publica.fraunhofer.de/handle/publica/428352This master thesis shows an advantage in using variational quantum circuits in a reinforcement learning algorithm to solve the Frozen Lake environment. It also looks into expressibility, entanglement and the effective dimension of the quantum circuits used, but no correlations are found between these operational descriptors and the performance of the hybrid quantum-classical reinforcement learning algorithm.enquantumreinforcement learningvariational quantum circuitexpressibilityentanglementeffective dimensionCharacteristics of Quantum Architectures in Reinforcement Learning Applicationsmaster thesis