Skip to main content
. 2022 Feb 28;18(2):e1009882. doi: 10.1371/journal.pcbi.1009882

Fig 5. Based on the ratios in the populations, SL-GA, SL-EC-Conf-Unc, SL-QL and SL-EC-Conf are the most dominating meta-social learning strategies in wide range of environments.

Fig 5

On the right of each figure, the ranks of the algorithms (higher to lower in terms of ratios) at the end of the processes are shown. The strategies were labeled and color coded to correspond to names provided on top. The differences that are not statistically significant (at p > 0.05, Wilcoxon rank-sum test) are linked by black vertical lines. The points of environment change are indicated with orange arrows. The highlighted areas show the standard deviation 112 runs.