A Novel Behavioral Strategy for RoboCode Platform Based on Deep Q-Learning

Table 1

The T-test results between the AI robot and Corners and Wall robots.

Corners vs. AI robot		Walls vs. AI robot
t Stat	−10.8663	t Stat	−7.4899
P (T ≤ t) one tail	5.73E−05	P (T ≤ t) one tail	6.92E−05
t Critical one-tail	2.015048	t Critical one-tail	1.894579
P (T ≤ t) two tail	0.000115	P (T ≤ t) two tail	0.000138
t Critical one-tail	2.570582	t Critical one-tail	2.364624