Machine Learning of Dose-Volume Histogram Parameters Predicting Overall Survival in Patients with Cervical Cancer Treated with Definitive Radiotherapy

Figure 3

The important factors selected by elastic-net models including all factors, which constructed in 5-fold cross validation and 100 bootstrapping iterations. (a) Clinical factors; (b) DVH parameters of OARs; (c) DVH parameters of tumor. X-axis is the mean coefficient of one factor, color is its value, and size is its frequency in 100 iterations.