Fig. 13
From: Evolution of cooperation guided by the coexistence of imitation learning and reinforcement learning
The MFC varies with introspection rate \(\alpha\) for three games as iterations proceed in a \(100 \times 100\) square lattice network: (a) CoG with \(a_{11}=3\), \(a_{12}=0\), \(a_{21}=0\) and \(a_{22}=2\); (b) CG with \(a_{11}=2\), \(a_{12}=0\), \(a_{21}=4\) and \(a_{22}=-1\); (c) PDG with \(a_{11}=-1\), \(a_{12}=-10\), \(a_{21}=0\) and \(a_{22}=-8\). The blue solid lines, red solid lines, green solid lines, and black solid lines show the proposed processes, with \(\alpha =0\), \(\alpha =0.3\), \(\alpha =0.7\), and \(\alpha =1\), respectively.
