Skip to main content

Table 5 Prediction accuracies of five runs of HBSA-KNN-based ensemble classifier on six test sets

From: Finding minimum gene subsets with heuristic breadth-first search algorithm for robust tumor classification

Run First Second Third Fourth Fifth Average
Dataset Acc.% Acc.% Acc.% Acc.% Acc.% Acc.%
Leukemia 86.54 84.62 88.46 84.62 84.62 85.57 ± 1.66
DLBCL 90.48 90.48 90.48 85.71 90.48 89.53 ± 2.13
Prostate 85.29 82.35 85.29 85.29 85.29 84.70 ± 1.31
SRBCT 95 95 95 90 95 94 ± 2.24
ALL 95 97 96 95 95 95.60 ± 0.89
Colon 75 75 75 75 75 75 ± 0
  1. Column First denotes the prediction accuracy of the constructed ensemble classifier obtained on the first run of the HBSA-KNN, and the others are deduced by analogy. The average accuracy is the average prediction accuracy obtained by five runs of the HBSA-KNN.