Figure 6.
(a) The best multivariate CPH c-index for each number of combinations trained with the entire dataset. (b) Comparison of the performance of DeepSurv, RSF, and CPH model in terms of survival probability calculated for an example patient. The dashed line represents the actual event time. (c) Comparing the statistics of the c-index for the train set as a function of the number of model features evidencing the median, first and third quartiles, and upper and lower bounds. (d) Same as the previous case but considering the test set.