Skip to main content

Table 3 Comparison of RBFNN performance on randomized datasets in 100 10xCV tests§

From: Identification of recent cases of hepatitis C virus infection using physical-chemical properties of hypervariable region 1 and a radial basis function neural network classifier

Dataset

No. CV runs

CA

F1 measure

MCC

AUROC

Train set 1a

1000

94.943% (±1.067)

0.960 (±0.009)

0.892 (±0.023)

0.981 (±0.005)

Train set 2

1000

95.958% (±0.717)

0.974(±0.005)

0.887 (±0.020)

0.986 (±0.003)

Train set 3

1000

96.014% (±0.719)

0.974 (±0.005)

0.889 (±0.020)

0.987 (±0.003)

Train set 4

1000

95.981% (±0.699)

0.974 (±0.005)

0.889 (±0.019)

0.986 (±0.003)

  1. §Comparisons are based on the corrected two-tailed T-test at a significance level of p < 0.001
  2. aDataset used to train (fit) the RBFNN classifier (1st and 2nd rows in Table 2)
Back to article page

BMC Genomics

ISSN: 1471-2164

Contact us