Research Article
New Dandelion Algorithm Optimizes Extreme Learning Machine for Biomedical Classification Problems
Table 6
Results comparisons between DA-ELM and fusion classifier for biomedical classification.
| Datasets | Algorithms | Training | Testing | Rate (%) | Dev | Rate (%) | Dev |
| EEG | DA-ELM | 69.78 | 0.0052 | 70.22 | 0.0062 | Max-ELM | 70.13 | 0.005 | 70.58 | 0.0053 | Min-ELM | 68.97 | 0.0087 | 69.72 | 0.0062 | Med-ELM | 69.93 | 0.0063 | 70.42 | 0.0059 | MV-ELM | 69.14 | 0.0091 | 69.13 | 0.0056 |
| Blood | DA-ELM | 79.81 | 0.0140 | 81.68 | 0.0133 | Max-ELM | 81.63 | 0.0138 | 81.96 | 0.0125 | Min-ELM | 80.56 | 0.0139 | 81.73 | 0.0142 | Med-ELM | 80.79 | 0.0142 | 81.81 | 0.0131 | MV-ELM | 79.06 | 0.0168 | 81.04 | 0.0125 |
| Statlog | DA-ELM | 86.22 | 0.0216 | 88.15 | 0.0175 | Max-ELM | 87.16 | 0.0208 | 89.95 | 0.0137 | Min-ELM | 86.56 | 0.0213 | 88.62 | 0.0142 | Med-ELM | 86.75 | 0.0218 | 88.16 | 0.0151 | MV-ELM | 88.74 | 0.0209 | 87.56 | 0.0143 |
| SPECT | DA-ELM | 81.35 | 0.0255 | 85.22 | 0.0243 | Max-ELM | 81.56 | 0.0209 | 86.83 | 0.0226 | Min-ELM | 81.73 | 0.0226 | 86.52 | 0.0235 | Med-ELM | 81.25 | 0.0237 | 85.69 | 0.0239 | MV-ELM | 80.68 | 0.0213 | 84.48 | 0.0232 |
|
|