Cross-Tissue Analysis Using Machine Learning to Identify Novel Biomarkers for Knee Osteoarthritis

Figure 4

In the least absolute shrinkage and selection operator (LASSO) regression model, when adjusted to the optimal lambda, the number of genes with a nonzero coefficient was 21 (a). The results of support vector machine recursive feature elimination (SVM-RFE) algorithm showed that when the number of genes was 28, the cross-validation error was smallest (b). The intersecting genes of the two methods included 14 genes (c).

(a)

(b)

(c)