Table 2: Classification of the performance of algorithms on average of more than 20 sections (the number in parentheses represents the standard error).
|
Datasets |
Methods |
Training dataset |
|
|
|
|
Testing dataset |
|
|
|
# selected descriptors |
AC |
SE |
SP |
MCC |
AC |
|
Breast
|
MI-BBA_NB |
13.65 (0.2497) |
0.9653 (0.0031) |
0.9878 (0.0029) |
0.9852 (0.0048) |
0.9562 (0.0056) |
0.9563 (0.0252) |
|
|
BBA |
494.95 (0.2528) |
0.8317 (0.0307) |
0.8507 (0.0285) |
0.7856 (0.0549) |
0.6156 (0.0729) |
0.9143 (0.0383) |
|
|
BGA |
401.85 (0.3108) |
0.9131 (0.0870) |
0.9753 (0.0872) |
0.8847 (0.1437) |
0.8826 (0.1870) |
0.9175 (0.0285) |
|
Leukemia
|
MI-BBA_NA |
14.3 (0.2693) |
0.9563 (0.0222) |
0.9856 (0.0861) |
0.9745 (0.0259) |
0.9632 (0.0041) |
0.9489 (0.0096) |
|
|
BBA |
465.95 (0.3086) |
0.9000 (0.0789) |
0.9222 (0.0750) |
0.8682 (0.1496) |
0.8830 (0.1749) |
0.9116 (0.0506) |
|
|
BGA |
397.05 (0.3621) |
0.9289 (0.0821) |
0.9289 (0.0735) |
0.8940 (0.1862) |
0.8527 (0.2086) |
0.8949 (0.0471) |
|
wisconsin
|
MI-BBA_NB |
14.2 (0.2608) |
0.9995 (0.0017) |
0.9925 (0.0154) |
0.9992 (0.0058) |
0.9989 (0.0036) |
0.9995 (0.0024) |
|
|
BBA |
474 (0.2721) |
0.9324 (0.0180) |
0.9306 (0.0299) |
0.9338 (0.0199) |
0.8544 (0.0402) |
0.9397 (0.0190) |
|
|
BGA |
383.55 (0.3021) |
0.9548 (0.248) |
0.9658 (0.0258) |
0.9320 (0.0403) |
0.8992 (0.0541) |
0.9366 (0.0096) |
By comparing the previous implementation between BBA and BGA with the proposed algorithm MI-BBA_NB, it shows us that it has a great ability in terms of accuracy of classification and also efficiency by applying them to four groups of bigdata.