Figure 3.

Example of the experimental set up optimization scheme applied to the wheat dataset 1. From the base population, a random sample of 50 genotypes was selected to build the test set. The remaining genotypes formed the candidates set were the optimization process takes places. From these candidates, four different algorithms using STPGA were used to build the training population set (TRS) test population size under two scenarios (Untargeted-Opt and Targeted-Opt). At the same time, a random sample of the same size was selected as the TRS. This process was repeated 100 runs. The entire set was repeated 30 times. Prediction accuracies (r) were calculated using GBLUP method.