ATTENTION/WARNING - NE PAS DÉPOSER ICI/DO NOT SUBMIT HERE

Ceci est la version de TEST de DIAL.mem. Veuillez ne pas soumettre votre mémoire sur ce site mais bien à l'URL suivante: 'https://thesis.dial.uclouvain.be'.
This is the TEST version of DIAL.mem. Please use the following URL to submit your master thesis: 'https://thesis.dial.uclouvain.be'.
 

Use of Machine Learning algorithm for prediction for survival data

Gengler, Rémi
(2023)

Files

Gengler_34161700_2023.pdf
  • Closed access
  • Adobe PDF
  • 1.9 MB

Details

Supervisors
Legrand, Catherine
Faculty
Faculté des sciences
Degree label
Master [120] en science des données, orientation statistique, à finalité spécialisée
Abstract
The Cox Proportional Hazards model is a popular model in Survival Analysis, dealing with the censoring case. In parallel, the use of Machine Learning, and especially of Artificial Neural Networks, has exploded in recent decades. Nevertheless, Artificial Neural Networks are not designed in themselves to model survival data: in particular, Artificial Neural Networks are not appropriate to handle censoring. However, several models tackling censoring and using Artificial Neural Networks have been proposed. Through this work, we present a review of the literature and in particular consider three models: models respectively proposed in Brown et al. 1997, Lisboa et al. 2003 and Katzman et al. 2018. The performances of these models are compared to the ones of the Cox Proportional Hazards model in two specific cases: for a very small number of training observations at hand and when the training observations contain ties in the survival times.