Survival analysis and neural nets

摘要: We consider feed-forward neural nets and their relation to regression models for survival data. show how the back-propagation algorithm may be used obtain maximum likelihood estimates in certain standard data, as well various generalizations of these. Examples concerning malignant melanoma post-partum amenorrhoea during lactation are illustration. conclude that although problems with substantial number parameters interpretation remain, network flexible extensions thereby candidates use prediction exploratory analyses larger data sets.