Fig. 10
From: Constructing phylogenetic networks via cherry picking and machine learning
Comparison of ML, TrivialRand, Hybroscale, and TreeChild on real data. Each point on the horizontal axis corresponds to one instance group, consisting of 10 instances for a fixed pair \((L,|\mathcal {T}|)\). In the top graph, the height of each bar gives the average, over all instances of the group, of the number of reticulations returned by the method. The bottom graphs compare the average output of ML within each instance group and the average of the best output given by TrivialRand within the group. The shaded areas represent 95% confidence intervals