Fig. 4From: Towards more efficient and robust evaluation of sepsis treatment with deep reinforcement learningComparison of learned strategies and the clinician strategyBack to article page