Summary of On Training Survival Models with Scoring Rules, by Philipp Kopper et al.
On Training Survival Models with Scoring Rules
by Philipp Kopper, David Rügamer, Raphael Sonabend, Bernd Bischl, Andreas Bender
First submitted to arxiv on: 19 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computation (stat.CO); Machine Learning (stat.ML)
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 A novel framework for training survival models is introduced, which leverages scoring rules to accommodate censoring in the context of survival analysis. The proposed method is model-agnostic, allowing for parametric or non-parametric learning of event-time distributions. Neural network-based implementations are showcased, along with proof-of-concept examples using gradient boosting, generalized additive models, and trees. Empirical evaluations on synthetic and real-world datasets demonstrate that scoring rules can be effectively incorporated into model training, achieving competitive predictive performance compared to established time-to-event models. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary In a breakthrough study, researchers have developed a new way to train survival models that uses scoring rules to help with censoring in survival analysis. This approach is flexible and can work with different types of models and data sets. The team tested their method on both made-up and real-world data and found that it performed well compared to other popular methods. |
Keywords
* Artificial intelligence * Boosting * Neural network
