Table 7. mycoSORT Results - Set of Features F5.
Under-sampling(USF) | Classifier | Precision | Recall | F-measure | MCC | F-2 |
Training set with USF 0% | Naive Bayes | 0.307 | 0.720 | 0.430 | 0.382 | 0.570 |
Training set with USF 0% | LMT | 0.656 | 0.420 | 0.512 | 0.485 | 0.450 |
Training set with USF 0% | LibSVM | 0.833 | 0.033 | 0.064 | 0.155 | 0.040 |
Training set with USF 5% | Naive Bayes | 0.310 | 0.733 | 0.436 | 0.390 | 0.580 |
Training set with USF 5% | LMT | 0.600 | 0.500 | 0.545 | 0.503 | 0.520 |
Training set with USF 5% | LibSVM | 0.703 | 0.173 | 0.278 | 0.319 | 0.200 |
Training set with USF 10% | Naive Bayes | 0.307 | 0.760 | 0.438 | 0.396 | 0.590 |
Training set with USF 10% | LMT | 0.574 | 0.567 | 0.570 | 0.523 | 0.570 |
Training set with USF 10% | LibSVM | 0.704 | 0.333 | 0.452 | 0.449 | 0.370 |
Training set with USF 15% | Naive Bayes | 0.309 | 0.793 | 0.445 | 0.41 | 0.600 |
Training set with USF 15% | LMT | 0.458 | 0.693 | 0.552 | 0.504 | 0.630 |
Training set with USF 15% | LibSVM | 0.596 | 0.413 | 0.488 | 0.451 | 0.440 |
Training set with USF 20% | Naive Bayes | 0.314 | 0.793 | 0.450 | 0.415 | 0.610 |
Training set with USF 20% | LMT | 0.422 | 0.653 | 0.513 | 0.460 | 0.590 |
Training set with USF 20% | LibSVM | 0.545 | 0.527 | 0.536 | 0.485 | 0.530 |
Training set with USF 25% | Naive Bayes | 0.312 | 0.780 | 0.446 | 0.408 | 0.600 |
Training set with USF 25% | LMT | 0.399 | 0.673 | 0.501 | 0.449 | 0.590 |
Training set with USF 25% | LibSVM | 0.481 | 0.580 | 0.526 | 0.470 | 0.560 |
Training set with USF 30% | Naive Bayes | 0.288 | 0.767 | 0.418 | 0.377 | 0.580 |
Training set with USF 30% | LMT | 0.388 | 0.727 | 0.506 | 0.461 | 0.620 |
Training set with USF 30% | LibSVM | 0.460 | 0.687 | 0.551 | 0.503 | 0.630 |
Training set with USF 35% | Naive Bayes | 0.302 | 0.780 | 0.435 | 0.397 | 0.590 |
Training set with USF 35% | LMT | 0.359 | 0.807 | 0.497 | 0.465 | 0.650 |
Training set with USF 35% | LibSVM | 0.369 | 0.800 | 0.505 | 0.472 | 0.650 |
Training set with USF 40% | Naive Bayes | 0.303 | 0.773 | 0.435 | 0.396 | 0.590 |
Training set with USF 40% | LMT | 0.344 | 0.840 | 0.488 | 0.463 | 0.650 |
Training set with USF 40% | LibSVM | 0.338 | 0.840 | 0.482 | 0.456 | 0.650 |
Results of Positive Class on Feature Setting #3, Using Only Bag-of-Words as Features.