Table 4 The table lists the average area (±SD) covered by the base classifiers across the 100-fold Monte Carlo cross-validation. The lowest area per dataset is highlighted in bold. The DT classifier has the lowest area for most of the datasets, i.e., the predictions are more stable. Refer to Fig. 4 (bottom) for the example showing the avp_amppred dataset

acp_mlacp

3.36 (±0.099)

2.93 (±0.084)

2.81 (±0.078)

2.8 (±0.076)

2.81 (±0.101)

aip_antiinflam

2.82 (±0.077)

2.48 (±0.051)

2.53 (±0.043)

2.34 (±0.076)

2.33 (±0.059)

amp_antibp2

3.09 (±0.087)

2.71 (±0.076)

3.16 (±0.129)

2.93 (±0.131)

2.82 (±0.091)

atb_antitbp

3.42 (±0.107)

3.2 (±0.087)

3.44 (±0.124)

3.31 (±0.106)

3.09 (±0.105)

avp_amppred

3.06 (±0.087)

2.6 (±0.065)

3.21 (±0.09)

2.98 (±0.168)

2.6 (±0.074)

cpp_mlcpp-complete

2.99 (±0.099)

2.55 (±0.081)

2.57 (±0.063)

2.48 (±0.078)

2.59 (±0.097)

hem_hemopi

3.3 (±0.073)

2.86 (±0.145)

3.16 (±0.105)

2.87 (±0.143)

2.86 (±0.156)

isp_il10pred

3.08 (±0.089)

2.72 (±0.055)

2.49 (±0.041)

2.53 (±0.064)

2.61 (±0.069)

nep_neuropipred

3.28 (±0.153)

2.81 (±0.096)

3.26 (±0.288)

3.16 (±0.347)

2.84 (±0.122)

pip_pipel

3.21 (±0.08)

2.36 (±0.044)

2.3 (±0.031)

2.33 (±0.048)

2.39 (±0.051)