Table 4 List of classification experiments
From: Feature analysis for classification of trace fluorescent labeled protein crystallization images
Exp ID | Tasks | No. of experiments 3 |
|---|---|---|
1 | Run all classifiers for 511 feature set (5 classifiers with/without normalization) | 2 * 5 * 511* 1 = 5110 |
2 | Run the best classifier 5 times and take the average for the best 70-feature set (RF) | 1 * 1 * 64 * 5 = 320 |
3 | Run classifiers PCA for 10,20,..,50 features | 1 * 5 * 5 * 2= 50 |
4 | Run classifiers using RF feature selection (10,20,...,50) | 1 * 5 * 5* 2 = 50 |
5 | Run BYS, DT and RF (with and without normalization, with graph features) for crystal sub categories | 2 * 3 * 511 * 1 = 3066 |
6 | Run RF, DT and BYS classifiers with and without normalization for likely-lead subcategories | 2 * 3 * 1 * 1 = 6 |
7 | Run RF, DT and BYS classifiers with and without normalization for non-crystal subcategories | 2 * 3 * 1 * 1 = 6 |
8 | Calculate training and testing time of the random forest for the largest feature | 1 * 1 * 1 * 5 = 5 |
9 | Calculate timings for feature extraction of an image | 1 * 1 * 11 *1 = 11 |
| Â | Total number of experiments | 8624 |