Multiagent Reinforcement Learning with Regret Matching for Robot Soccer

<table>Scores difference <svg height="11.175" id="M218" style="vertical-align:-0.0pt" version="1.1" viewbox="0 0 13.775 11.175" width="13.775" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
<g transform="matrix(.017,-0,0,-.017,.062,11.113)"><path d="M780 372q0 -191 -178 -297q-128 -75 -335 -75h-250l7 27q61 7 77 22t28 76l77 398q11 62 0 76.5t-77 22.5l6 28h292q164 0 258.5 -67.5t94.5 -210.5zM678 375q0 119 -73 180.5t-201 61.5q-62 0 -82 -13q-17 -7 -25 -53l-80 -427q-8 -47 -4 -62t26 -20q24 -6 80 -6
q166 0 262.5 95t96.5 244z" id="x1D437"></path></g>
</svg> as the number of matches increases.</table>

Mathematical Problems in Engineering

fig6

Figure 6

Figure 6: Multiagent Reinforcement Learning with Regret Matching for Robot Soccer 