Selection and Penalty Strategies for Genetic Algorithms Designed to Solve Spatial Forest Planning Problems
Figure 6
Results for Landscape C, 6000 generations. This figure presents a box-plot of solution
quality (NPV; -axis) for each paired selection/penalty strategy tested. The -axis displays the strategy; static and dynamic penalty strategies for the same
selection strategy are displayed side by side (static on the left side). Reading from left
to right, the first selection strategy presented is proportional selection and the rest are
deterministic tournaments with various competitor pool sizes (). The label “” corresponds to the strategy wherein was sampled from a uniform distribution over the range (2, 10), and the label
“” corresponds to the strategy where increased from 2 to 7 during the search.
