Table 9 Comparison of our proposed BBA based feature selection methodology with various well-known meta-heuristic feature selection algorithms used in the literature. (Here, bold indicates the least number of features selected and highest values of accuracy, Recall, Precision and F1-score attained for each of the HAR datasets).
Dataset | Optimization algorithm Used | Number of features selected | Accuracy (%) | Recall | Precision | F1-Score |
|---|---|---|---|---|---|---|
HARTH | Cuckoo Search Algorithm(CSA)50 | 905 | 87.3 | 0.873 | 0.889 | 0.873 |
Equilibrium Optimize(EO)51 | 1105 | 88.29 | 0.883 | 0.891 | 0.888 | |
Genetic Algorithm(GA)52 | 1066 | 88.48 | 0.885 | 0.894 | 0.883 | |
Gravitational Search Algorithm (GSA)53 | 1135 | 87.48 | 0.875 | 0.894 | 0.883 | |
Grey Wolf Optimizer(GWO)54 | 1346 | 88.49 | 0.885 | 0.891 | 0.888 | |
Proposed Model (BBA) | 906 | 88.89 | 0.889 | 0.901 | 0.888 | |
KU-HAR | Cuckoo Search Algorithm(CSA)50 | 1208 | 97.54 | 0.975 | 0.956 | 0.955 |
Equilibrium Optimize(EO)51 | 1125 | 97.76 | 0.955 | 0.958 | 0.958 | |
Genetic Algorithm(GA)52 | 1157 | 97.65 | 0.956 | 0.957 | 0.958 | |
Gravitational Search Algorithm (GSA)53 | 1106 | 97.86 | 0.959 | 0.959 | 0.959 | |
Grey Wolf Optimizer(GWO)54 | 1470 | 97.76 | 0.958 | 0.958 | 0.957 | |
Proposed Model (BBA) | 1049 | 97.97 | 0.965 | 0.965 | 0.965 | |
HuGaDB | Cuckoo Search Algorithm(CSA)50 | 1113 | 91.57 | 0.916 | 0.922 | 0.916 |
Equilibrium Optimize(EO)51 | 1100 | 91.01 | 0.910 | 0.916 | 0.911 | |
Genetic Algorithm(GA)52 | 1047 | 91.57 | 0.916 | 0.922 | 0.916 | |
Gravitational Search Algorithm(GSA)53 | 1122 | 91.01 | 0.910 | 0.916 | 0.911 | |
Grey Wolf Optimizer(GWO)54 | 1423 | 91.01 | 0.910 | 0.915 | 0.912 | |
Proposed Model (BBA) | 1017 | 93.82 | 0.939 | 0.943 | 0.939 |
