Open Access

Table 4

Performances of all classifiers.

No. Classifiers Data type Training Test Average test AUc
Sensitivity Specificity Accuracy Sensitivity Specificity Accuracy
1 KNN CBC readings 98.1% 98.3% 98.2% 80.7% 82.6% 81.7% 0.901
2 Discriminant Analysis CBC readings 76.0% 82.0% 79.0% 70.1% 85.1% 77.6% 0.854
3 SVM CBC readings 77.4% 81.6% 79.5% 73.4% 85.7% 79.6% 0.857
4 Naive Bayes CBC readings 79.9% 81.4% 80.6% 82.2% 80.1% 81.1% 0.860
5 Decision Tree CBC readings 82.5% 81.2% 81.8% 73.4% 75.4% 74.4% 0.832
6 Logistic Regression CBC readings 76.6% 81.9% 79.2% 77.3% 83.3% 80.3% 0.866
7 Random Forests CBC readings 98.2% 98.2% 98.2% 82.7% 82.1% 82.4% 0.899
8 ANN CBC readings 81.3% 84.4% 82.8% 78.7% 83.0% 80.9% 0.881
9 CNN1 SFL-SSC images 97.9% 98.9% 98.6% 93.9% 96.6% 95.6% 0.981

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.