Skip to main content

Table 7 Classification performance of parsimony models over four disease gene-expression data using four information theoretic filters

From: iRDA: a new filter towards predictive, stable, and enriched candidate genes

  mRMR CMIM FCBF iRDA
Error % # % # % # % #
Multiple Myeloma (MM) 14.45 7 14.45 6 15.03 10 16.18 10
Marfan Syndrome (MS) 3.96 4 6.93 7 11.88 3 5.94 6
HIV Infection (HIV) 13.86 8 11.45 8 15.06 7 10.84 8
Neurodegeneration (AD) 0.84 2 0 2 0.84 2 0 2
avg 8.28 5.25 8.21 5.75 10.7 5.5 8.24 6.5
AUC % # % # % # % #
Multiple Myeloma (MM) 90.71 9 90.09 9 91.26 10 89.84 9
Marfan Syndrome (MS) 98.84 4 97.2 3 93.25 7 98.76 4
HIV Infection (HIV) 91.88 8 92.42 8 87.29 7 92.49 8
Neurodegeneration (AD) 100 2 100 2 100 2 100 2
avg 95.36 5.75 94.93 5.5 92.95 6.5 95.27 5.75
MCC % # % # % # % #
Multiple Myeloma (MM) 61.52 7 63.67 6 61.55 10 59.56 10
Marfan Syndrome (MS) 91.93 4 85.7 7 78.67 6 88.61 6
HIV Infection (HIV) 63.06 8 68.98 8 57.99 7 70.4 8
Neurodegeneration (AD) 98.33 2 100 2 98.33 2 100 2
avg 78.71 5.25 79.59 5.75 74.14 6.25 79.64 6.5
  1. %: performance rate; #: number of explored genes
\