Skip to main content
. 2020 Oct 22;10:18074. doi: 10.1038/s41598-020-75029-1

Table 3.

Comparison among the best subsets found and top-k ranked descriptors.

Datasets Instances Best merit scoresa Merit scores of top-k ranked descriptorsb
k
20 30 40 50 60
Overall_NR98 32300 3.985 3.92 3.96 3.962 3.958 3.942
Overall_NR90 22512 3.87 3.79 3.827 3.829 3.822 3.807
Overall_NR70 14559 3.667 3.577 3.608 3.616 3.607 3.589
Overall_NR50 9428 3.474 3.368 3.396 3.406 3.403 3.388
Overall_NR30 4735 3.13 3.051 3.057 3.044 3.038 3.035
Antibacterial_NR98 10303 3.41 3.336 3.361 3.364 3.356 3.334
Antifungal_NR98 4546 3.059 2.98 3.018 3.01 3.011 2.993
Antiviral_NR98 3849 3.031 2.956 2.992 3.001 3.001 2.995
Anticancer_NR98 1557 2.588 2.517 2.552 2.55 2.54 2.529
Antiparasitic_NR98 501 2.11 2.042 2.063 2.06 2.065 2.061
Averages 3.232 3.154 3.183 3.184 3.180 3.167

aMerit scores of best subsets found.

bMerit scores of the top-k ranked descriptors from the candidate feature set.