Skip to main content

Table 3 Confusion Matrices.

From: Biomarker Discovery and Redundancy Reduction towards Classification using a Multi-factorial MALDI-TOF MS T2DM Mouse Model Dataset

     CHF   HF   SD     
nFeat   Method   CHF HF SD   CHF HF SD   CHF HF SD   Error   P-Value
3   ANOVA   33 14 18   17 24 24   16 16 36   0.53   0.0028
   ACO   45 4 16   10 33 22   11 16 41   0.4   1e-08
   Cluster ANOVA   40 12 13   15 28 22   9 16 43   0.44   1e-06
5   ANOVA   36 13 16   18 22 25   20 11 37   0.52   0.006
   ACO   48 3 14   12 33 20   10 15 43   0.37   2.7e-08
   Cluster ANOVA   40 13 12   15 38 12   4 24 40   0.4   6.7e-07
8   ANOVA   41 12 12   16 34 15   6 22 40   0.42   9e-06
   ACO   45 5 15   12 30 23   4 18 46   0.39   5.5e-07
   Cluster ANOVA   43 10 12   14 35 16   5 19 44   0.38   3.3e-07
  1. Confusion matrix for 10-fold cross validation for experimental factor diet using random forest classifier. The feature selection was done by three different methods: ANOVA, ant colony optimization (ACO) and cluster-based ANOVA. The feature selection was performed three times with different number of features: 3, 5 and 8. Numbers in bold print represents true positives.