Skip to main content

Table 1 AUC performance for ADNI and SSI data across different ratio of training and testing datasets obtained by 10-fold cross-validation

From: DL-CHI: a dictionary learning-based contemporaneous health index for degenerative disease monitoring

α=0, β    
Data Ratio (%) CHI DL-CHI
ADNI 15 0.870 ± 0.024 0.887 ± 0.021
  20 0.883 ± 0.021 0.890 ± 0.016
  35 0.889 ± 0.014 0.936 ± 0.051
  50 0.890 ± 0.031 0.940 ± 0.047
  75 0.927 ± 0.012 0.959 ± 0.036
SSI 15 0.850 ± 0.055 0.867 ± 0.039
  20 0.861 ± 0.036 0.877 ± 0.020
  35 0.871 ± 0.012 0.886 ± 0.020
  50 0.862 ± 0.015 0.892 ± 0.041
  75 0.889 ± 0.024 0.914 ± 0.027
α, β=0    
ADNI 15 0.780 ± 0.016 0.863 ± 0.034
  20 0.799 ± 0.054 0.873 ± 0.024
  35 0.804 ± 0.012 0.844 ± 0.034
  50 0.818 ± 0.019 0.869 ± 0.064
  75 0.855 ± 0.064 0.905 ± 0.024
SSI 15 0.829 ± 0.064 0.860 ± 0.023
  20 0.860 ± 0.021 0.879 ± 0.016
  35 0.870 ± 0.034 0.883 ± 0.034
  50 0.880 ± 0.042 0.892 ± 0.036
  75 0.883 ± 0.026 0.895 ± 0.016
α, β    
ADNI 15 0.865 ± 0.021 0.872 ± 0.025
  20 0.871 ± 0.023 0.881 ± 0.014
  35 0.874 ± 0.032 0.890 ± 0.026
  50 0.891 ± 0.021 0.910 ± 0.041
  75 0.901 ± 0.020 0.919 ± 0.036
SSI 15 0.741 ± 0.032 0.814 ± 0.041
  20 0.758 ± 0.034 0.820 ± 0.030
  35 0.770 ± 0.013 0.831 ± 0.036
  50 0.791 ± 0.026 0.887 ± 0.015
  75 0.806 ± 0.010 0.862 ± 0.036
\