Table 5 Epoch size along with the number of features selected before and after applying BBA based feature selection method for all the three HAR datasets.

From: Wrapper-based deep feature optimization for activity recognition in the wearable sensor networks of healthcare systems

Dataset

Model

#Runs with epoch

#Features before using BBA

#Features after using BBA

HARTH

Mobilenet_v3_large

40

960

471

Efficientnet_b0

40

1280

464

Concatenated feature vector

40

2240

906

KU-HAR

Mobilenet_v3_large

40

960

499

Efficientnet_b0

40

1280

721

Concatenated feature vector

40

2240

1249

HuGaDB

Mobilenet_v3_large

40

1280

621

Efficientnet_b0

40

960

423

Concatenated feature vector

40

2240

1067

Back to article page