Fig. 2: Performance of the Top-12 LASSO–SVM classifier.
From: Predictive modeling of chronic foot ulcer outcomes using longitudinal photoacoustic imaging
a Macro-averaged ROC curve for the Top-12 model; the blue solid line indicates the macro-averaged ROC and the gray dashed line indicates the diagonal reference (no-skill) classifier. b Averaged confusion matrix for the Top-12 feature model across 10 random train–test splits. The color scale from cyan to magenta represents the averaged classification accuracy (proportion of samples) in each cell, with magenta indicating higher values and cyan indicating lower values. Healing: n = 27; Worsening: n = 12; Healthy: n = 24.
