Learning from Demonstrations and Human Evaluative Feedbacks: Handling Sparsity and Imperfection Using Inverse Reinforcement Learning Approach

Figure 6

(a) The relation between sparsity level of demonstrations in stage-one and the number of feedbacks needed to reach “” score equal to 1.17 using . (b)’s stage-two performance in face of optimal and different demonstration sparsity levels in stage-one (point “B1”, …, “B10” in Figure 3) and the number of evaluative feedbacks. The black curve has no initial demonstration (point “C” in Figure 3).

(a)

(b)