Task assignment for hybrid scenarios in spatial crowdsourcing: A Q-Learning-based approach