Skip to main content

Table 2 RF modeling, filtered data, approach I

From: Data mining of high density genomic variant data for prediction of Alzheimer's disease risk

Data RF 10-Fold CV %   Error
(APOEε4+ & ε4- samples) Number of Trees F = 7 F = 11 F = 18 F = 32 F = 42
APOE & 10 GAB2 SNPs       
  100 33.3 33.2    
  300 33.3 33.2    
  600 33.4 33.5    
199 SNPs       
  100   11.9 12.3 12.5 12.3
  300   11.3 11.3 12.1 11.6
  600   10.6 11.6 11.2 11.1
APOE & 199 SNPs       
  100   10.9 9.9 10.4 10.3
  300   8.9 9.6 9.5 10.3
  600   9.0 9.1 10.1 9.9
APOE , 10 GAB2 , & 199 SNPs       
  100   9.9 10.1 9.9 11.2
  300   8.9 9.1 9.5 10.3
  600   9.1 9.4 9.9 10.3
  1. F = number of features.