Table 3 Comparative outcomes of BGWO-EDLMHAR technique with existing methods18,45,46.
Classifier | Accuy | Precn | Recal | F1score |
|---|---|---|---|---|
BGWO-EDLMHAR | 98.51 | 94.70 | 91.76 | 93.11 |
SHO-LSTM | 97.82 | 90.26 | 89.03 | 91.56 |
MFCC | 98.01 | 93.13 | 87.75 | 91.26 |
CA-WGNN | 96.68 | 93.29 | 88.37 | 92.48 |
RecurrentHAR | 96.26 | 89.19 | 89.43 | 87.72 |
DeepConvLG | 98.01 | 90.93 | 88.19 | 87.27 |
ResNet-BiGRU-SE | 98.11 | 89.77 | 88.63 | 91.54 |
Generic algorithm | 97.24 | 89.61 | 88.46 | 90.86 |
CNN-BiLSTM | 97.40 | 92.55 | 86.96 | 90.53 |
TAHAR-student-LSTM | 96.04 | 93.37 | 87.71 | 92.86 |
HMM method | 94.36 | 89.05 | 88.96 | 90.02 |
