Summary of An Introduction to Deep Survival Analysis Models For Predicting Time-to-event Outcomes, by George H. Chen
An Introduction to Deep Survival Analysis Models for Predicting Time-to-Event Outcomes
by George H. Chen
First submitted to arxiv on: 1 Oct 2024
Categories
- Main: Machine Learning (stat.ML)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper provides a modern introduction to survival analysis, focusing on predicting time-to-event outcomes at the individual data point level using neural networks. The goal is to understand the basic time-to-event prediction problem, its differences from standard regression and classification, and how key “design patterns” have been used to derive new models. The authors discuss classical methods like the Cox proportional hazards model and modern deep learning approaches such as deep kernel Kaplan-Meier estimators and neural ordinary differential equation models. They also extend the basic setup to predicting which of several critical events will happen first, along with the time until this earliest event happens (competing risks setting), and predicting time-to-event outcomes given a time series that grows in length over time (dynamic setting). The paper concludes with discussions on fairness, causal reasoning, interpretability, and statistical guarantees. An accompanying code repository implements every model and evaluation metric covered. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper is about using computers to predict when important events will happen. For example, it can help us know when a customer will stop paying for something or when someone will wake up from a coma. The authors use special kinds of computer programs called neural networks to make these predictions. They explain how this works and show examples of different ways to do it. They also talk about situations where there are multiple events that could happen, like knowing which event will happen first. This paper is important because it helps us understand how to use computers to make accurate predictions. |
Keywords
» Artificial intelligence » Classification » Deep learning » Regression » Time series
