Figure 3.

Machine learning analysis

(A) Dot plot showing the k-means clustering on latent space generated by autoencoder. Reprinted and Adapted from Wang et al.¹

(B) Bar plot showing the accuracy of classification in all 5-fold validation groups.

(C) Confusion matrix illustrating the predicted labels and true labels of samples classified by logistic classification.

(D) Receiver operator characteristic curves of prediction models trained on each individual omics datasets, combined omics (cytokines, proteomics, and metabolomics), and the minimal panel using molecular profile. Reprinted and Adapted from Wang et al.¹