Skip to main content

Advertisement

Table 8 ANN prediction based on 47 blinded samples in the cord blood data set

From: A simpler method of preprocessing MALDI-TOF MS data for differential biomarker analysis: stem cell and melanoma cancer studies

  50 ANN sub-models     
Sample label High Low ANN output Std err in 95% CI ANN classification Target output
Blind 1 48 2 0.040 0.057 High High
Blind 2 50 0 0 0 High High
Blind 3 50 0 0 0 High High
Blind 4 50 0 0 0 High High
Blind 5 50 0 0 0 High High
Blind 6 50 0 0 0 High High
Blind 7 35 15 0.300 0.132 High High
Blind 8 50 0 0 0 High High
Blind 9 48 2 0.040 0.057 High High
Blind 10 41 9 0.180 0.111 High High
Blind 11 48 2 0.040 0.057 High High
Blind 12 50 0 0 0 High High
Blind 13 40 10 0.200 0.115 High High
Blind 14 50 0 0 0 High High
Blind 15 46 4 0.080 0.078 High High
Blind 16 49 1 0.020 0.040 High High
Blind 17 50 0 0 0 High High
Blind 18 50 0 0 0 High High
Blind 19 41 9 0.180 0.111 High High
Blind 20 28 22 0.440 0.143 High High
Blind 21 50 0 0 0 High High
Blind 22 11 39 0.780 0.120 Low Low
Blind 23 0 50 1 0 Low Low
Blind 24 0 50 1 0 Low Low
Blind 25 2 48 0.960 0.057 Low Low
Blind 26 2 48 0.960 0.057 Low Low
Blind 27 0 50 1 0 Low Low
Blind 28 0 50 1 0 Low Low
Blind 29 0 50 1 0 Low Low
Blind 30 0 50 1 0 Low Low
Blind 31 0 50 1 0 Low Low
Blind 32 0 50 1 0 Low Low
Blind 33 0 50 1 0 Low Low
Blind 34 0 50 1 0 Low Low
Blind 35 2 48 0.960 0.057 Low Low
Blind 36 34 16 0.320 0.135 High* Low
Blind 37 19 31 0.620 0.140 Low Low
Blind 38 0 50 1 0 Low Low
Blind 39 5 45 0.900 0.087 Low Low
Blind 40 16 34 0.680 0.135 Low Low
Blind 41 0 50 1 0 Low Low
Blind 42 8 42 0.840 0.106 Low Low
Blind 43 5 45 0.900 0.087 Low Low
Blind 44 0 50 1 0 Low Low
Blind 45 0 50 1 0 Low Low
Blind 46 0 50 1 0 Low Low
Blind 47 0 50 1 0 Low Low
  1. High and low refer to the quantity of stem cells in cord blood. ANN output is computed based on the average performance from 50 random samplings. Std err with 95% CI refers to the standard error for ANN output in 95% confident interval range. ANN classification indicates the final outcome of the model. Target output refers to the original group to which the sample belongs to.