Reinforcement Learning And Optimal Experimental Design For Modeling And Control Of Fish Schooling Hydrodynamics