Research Article

Development of a Novel Deep Learning-Based Prediction Model for the Prognosis of Operable Cervical Cancer

Figure 5

Training and validation of DSLM.. (a) The values of the loss function for DSLM decrease from 20,000 to 11,794 after 2000 iterations. (b) The prediction error curve with IBS ()ā€‰=0.11 for DSLM. (c) Sequential values of actual and predicted number of patient deaths for each time interval in validation set. The red line indicates the actual number of patient deaths, while the blue line indicated the predicted number of patient deaths. The median absolute error equals to 7.712, while the mean absolute error equals to 8.612.
(a)
(b)
(c)