Research Article
Daily Human Physical Activity Recognition Based on Kernel Discriminant Analysis and Extreme Learning Machine
Table 2
Classification results of ELM, LM-BP, and LS-SVM classifiers.
| Features | Classifier | Time (s) | Accuracy rate (%) | Training | Testing | Training | Testing |
| Original | ELM | 0.0022 | 0.0031 | 96.44 | 96.07 | KDA based | ELM | 0.0031 | 0.0012 | 99.96 | 99.81 | Original | LM-BP | 1.8985 | 0.0175 | 95.59 | 95.24 | KDA based | LM-BP | 0.8861 | 0.0246 | 98.48 | 98.02 | Original | LS-SVM | 0.7775 | 0.1819 | 97.36 | 95.97 | KDA based | LS-SVM | 0.7678 | 0.2760 | 99.94 | 99.05 |
|
|