Reinforcement Learning target performance estimation on DR19-75-1-late