Skip to main content
Figure 3 | BMC Bioinformatics

Figure 3

From: An adaptive genetic algorithm for selection of blood-based biomarkers for prediction of Alzheimer's disease progression

Figure 3

Cross-validation parameters assessment. a) Mean time from each set of AGA runs, the AGA was run 100 times with each set of the parameters listed in table 2 respectively, Y axis is the average time of the 100 runs for each run ID. b) RMS errors between the AUC reported by AGA runs (100 for each run ID) and the 1000-repeated post validation (mean of the 1000 AUC). The number of repeats and number of fold for Run ID 1 to 7 are: 1*5, 3*5, 5*5, 3*10,5*10,10*5 and 15*5 respectively. For all these runs, other AGA parameters were: ρ = 0, no penalty to size of feature set; n max = 50, reduce computational impact; and skew = 1/6, random initial population skewed toward 1/6 features selected.

Back to article page