Fig 4. SL-GA and SL-EC-Conf-Unc show the best performance vs. exploration cost on diverse set of environments.

Overall, the meta-social learning strategies that utilize conformist strategy show better performance environments with high uncertainty. On the right of each figure, the color coded labels indicate the meta-social learning strategies (A through M maps to the strategy with the same color shown on top), and are ranked based on their performance from the best to worst. The decimal numerals on the left of the color codes indicate their average ranks (the lower is the better). The strategies are linked with vertical bold lines if the differences in their average ranks are not statistically significant at α = 0.05 based on the post-hoc Nemenyi test.