The Deep Promotion Time Cure Model

Victor Medina-Olivares, Stefan Lessmann, Nadja Klein
5/19/2023
stat.MLcs.LG

No official code found. Best implementation: vhmedina/deepcure (0 stars, MIT, last commit 2024-07-17).

Abstract

We propose a novel method for predicting time-to-event in the presence of cure fractions based on flexible survivals models integrated into a deep neural network framework. Our approach allows for non-linear relationships and high-dimensional interactions between covariates and survival and is suitable for large-scale applications. Furthermore, we allow the method to incorporate an identified predictor formed of an additive decomposition of interpretable linear and non-linear effects and add an orthogonalization layer to capture potential higher dimensional interactions. We demonstrate the usefulness and computational efficiency of our method via simulations and apply it to a large portfolio of US mortgage loans. Here, we find not only a better predictive performance of our framework but also a more realistic picture of covariate effects.

View on arXivView PDF

Code Implementations(1)

vhmedina/deepcureCommunity100%
MIT
01Python, Jupyter NotebookMay 16, 20232 years agoMIT

Similar Papers

An Introduction to Deep Survival Analysis Models for Predicting Time-to-Event Outcomes

85% match2 repos512024

Case-Base Neural Networks: survival analysis with time-varying, higher-order interactions

80% match2023

Towards Flexible Time-to-event Modeling: Optimizing Neural Networks via Rank Regression

80% match1 repo42023

Gradient-based Explanations for Deep Learning Survival Models

80% match1 repo42025

Deep Learning-Based Survival Analysis with Copula-Based Activation Functions for Multivariate Response Prediction

80% match2025

Flexible Deep Neural Networks for Partially Linear Survival Data

80% match2025

Cite this paper

@article{medinaolivares2023deep,
  title  = {The Deep Promotion Time Cure Model},
  author = {Victor Medina-Olivares and Stefan Lessmann and Nadja Klein},
  year   = {2023},
  eprint = {2305.11575},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2305.11575v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2305.11575.svg)](https://codeofpaper.com/paper/2305.11575)

Discussion