Evaluation of the dependence of radiomic features on the machine learning model.

Results for each dataset can be found in Additional file 5 . Predictive performance and feature stability Comparing the mean AUC-ROC to the number of statistically similar models showed a slightly significant and decreasing association ( p = 0.029; Fig. 5 a), that is, the better the best model performed, the less statistically similar models were observed.

also in 1,762 other papers