Table 3.
Comparison of software packages based on classification by clustering method. The accuracies of the machine learning models used for the homogeneous datasets are presented.
| Model | ODMS CA (%) | WDMS CA (%) | RMS CA (%) |
|---|---|---|---|
| Naive Bayes | 79.9 | 77.0 | 80.8 |
| Generalised Linear Model | NA | NA | 82.7 |
| Logistic Regression | 94.1 | 74 | 22.9 |
| Fast Large Margin | NA | NA | 83.3 |
| Deep Learning/Neural Network | 94.2 | NA | 86.1 |
| Decision Tree | 62.3 | 77.0 | NA |
| Random Forest | 73.9 | 83.0 | 55.1 |
| Stochastic Gradient Descent | 94.5 | 71.0 | 87.1 |
| Support Vector Machine | 94.0 | 75.0 | 78.3 |
| Average based on Available Models | 84.7 | 76.2 | 72.0 |
Legend: ODMS = Orange Data Mining Software, WDMS = Weka Data Mining Software and RMS = RapidMiner Studio (RMS), NA = Not available.