Skip to main content
. 2022 May 28;12(6):1344. doi: 10.3390/diagnostics12061344

Table 8.

Testing accuracy results using 50 most significant automatic features.

Image Categorization General Pattern Type Grading Grade Grading
Model 1 Model 2
1000 features 88.3% 93.9% 72.2% 83.9%
PCA 86.4% 91.2% 72.2% 81.2%
ECFS 75.9% 86.5% 68.4% 80.0%
ILFS 70.6% 79.3% 66.1% 78.5%
Fisher 74.6% 84.2% 65.7% 75.6%

PCA, principal component analysis; ECFS, ensemble-based classifier feature selection; ILFS, infinite latent feature selection.