Figure 5 | Scientific Reports

Figure 5

From: Intrinsic interactive reinforcement learning – Using error-related potentials for real world human-robot interaction

Figure 5

Simulated robot learning. (a) Accumulated errors of the robot for each subjects, (b) accumulated regret for each subjects, (c) accumulated errors of the robot in the first half of the experiment (1 to 45 actions) for each subject, (d) accumulated errors of the robot in the second half of the experiment (46 to 90 actions) for each subject, (e) total number of the robot’s errors in the first or the second half of the experiment for each subject, and (f) comparison between the first and last half of the experiment (1 to 45 actions vs. 46 to 90 actions) by performing Wilcoxon sign-rank test (two-sided, alpha = 0.05). The raw values (sample size of 7, i.e., 7 data pairs) used for this statistical analysis were depicted in (e).

Back to article page