Table 10.
Averaged and truncated confusion matrix for multi-class classification using the standard-sized test dataset, for 10 runs, using labels. From left to right, using 10/35/35, 20/40/40 and 30/35/35 percent of the labels for COVID-19, Pneumonia and normal diagnostics, respectively. From top to bottom, the supervised model, the SSDL model with no PBC, and the SSDL model with the PBC.
![]() |