Fig. 6
From: A dosing strategy model of deep deterministic policy gradient algorithm for sepsis patients
The differences in the sample size distribution of IV fluid dosage resulting from the use of the DDPG (part a in Fig. 6) and DQN (part b in Fig. 6) algorithms are plotted for a case in which the difference between the model and VP dosing decision of human clinicians is zero. As shown, the treatment plan generated by the model based on the DDPG algorithm is closer to the treatment plan generated by a human clinician