Abstract:
|
A new method of using artificial neural network to predict the survival probability function of competing risks will be presented. This new method of utilizing artificial neural network was trained using an updated version of the Bayesian Learning method that was introduced by Neal in 1996 and uses the hybrid Monte Carlo algorithm. A comparison was also structured to test the effectiveness of the updated Bayesian learning to the existing one. All of the developed models were applied to real data, taken from the Surveillance, Epidemiology, and End Results Program (SEER), which is conducted by the National Cancer Institute.
|