Page 130 - Data Science Algorithms in a Week

114 Loris Nanni, Sheryl Brahnam and Alessandra Lumini

Table 2 (Continued.)

WAVE Tr CW RS DCT FFT
hab 58.9 70.1 61.2 66.5 68.3
vote 60.1 96.9 82.6 96.7 97.8

aust 85.6 92.2 90.7 91.6 92.2
trans 62.1 71.2 64.7 69.9 71.5
wdbc 95.1 99.4 99.3 99.3 99.5
bCI 81.9 94.6 95.0 95.6 95.4

pap 72.4 80.7 84.0 85.2 85.5
torn 80.2 85.2 91.1 90.3 90.4
gCr 69.6 71.1 78.3 78.9 79.6
Average 75.5 86.3 84.9 87.1 87.6

Table 3: Performance (AUC) of the ensemble created by fixing the descriptor (first
four columns) and the reshaping method (last four columns).

DATASET DLPQ DCLBP DHoG DWave RTr RCW RRS RDCT RFFT
breast 97.5 97.9 99.5 99.4 99.2 99.2 99.3 99.3 99.2
heart 89.3 89.3 90.2 89.9 89.5 90.6 90.3 89.3 90.1
pima 72.0 72.2 80.8 82.3 74.4 80.9 80.8 80.6 80.5

sonar 92.8 89.3 94.2 92.6 70.9 93.9 93.1 93.7 93.6
iono 98.6 97.9 98.2 97.8 92.6 98.2 98.6 98.6 98.4

liver 71.8 70.4 73.4 73.4 59.3 73.2 74.2 73.4 73.6
hab 62.6 61.5 69.0 69.2 60.7 66.4 69.5 67.0 68.1
vote 97.8 97.3 98.1 96.8 74.7 97.1 98.3 97.6 97.8
aust 90.4 90.9 91.2 92.1 83.8 91.4 91.9 91.6 91.8

trans 68.3 67.1 69.2 71.0 66.0 67.1 69.9 68.7 70.1
wdbc 98.8 98.4 99.4 99.5 94.5 99.4 99.4 99.5 99.6

bCI 96.5 96.2 96.6 95.2 83.7 96.4 95.5 96.5 96.8
pap 86.8 82.4 87.0 84.3 74.4 84.9 86.6 87.4 87.6
torn 92.8 93.4 94.0 89.4 85.2 92.9 94.8 94.5 94.6
gCr 77.1 76.8 77.5 77.4 68.3 75.0 78.5 79.1 79.8

Average 86.2 85.4 87.9 87.4 78.5 87.1 88.0 87.8 88.1