Fig. 5

Development and evaluation of miRNAs-based machine learning cervical cancer survival prediction model. a Workflow of model development and evaluation. b The performance of classification for the prediction model was evaluated by ROC (Receiver operating characteristics) curve and AUC (Area under curve) value. c The accuracy of survival subtype prediction was assessed by Kaplan–meier analysis and Log-rank p value. SVM algorithm implementation was performed using svm() function in “e1071” (version 1.7–6) package and predict() function in “car” (version 3.0–10). ROC curve analysis was performed using roc() function in “pROC” (version 1.17.0.1) package and plot() function in “graphics” (version 3.6.1). Kaplan–meier analysis was performed using Surv() and survfit() function in “survival” (version 3.2–10) package. The Kaplan–meier plot was performed using ggsurvplot() function in “survminer” (version 0.4.9). RStudio (version 3.6.1, RStudio, Inc.) is used