Summary of Teaching Models to Survive: Proper Scoring Rule and Stochastic Optimization with Competing Risks, by Julie Alberge (soda) et al.
Teaching Models To Survive: Proper Scoring Rule and Stochastic Optimization with Competing Risks
by Julie Alberge, Vincent Maladière, Olivier Grisel, Judith Abécassis, Gaël Varoquaux
First submitted to arxiv on: 20 Jun 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: None
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 proposes a novel approach for predicting multiple outcomes in right-censored data, where some events are missing due to limited observation periods. This problem is known as competing risks. To address this challenge, the authors introduce a strictly proper censoring-adjusted separable scoring rule that can be optimized on a subpart of the data independently of observations. This allows for stochastic optimization using gradient boosting trees. The proposed model, MultiIncidence, outperforms 11 state-of-the-art models in estimating outcome probabilities in survival and competing risks tasks. Notably, MultiIncidence can predict outcomes at any time horizon and is significantly faster than existing alternatives. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps computers better understand when some events don’t happen because we can’t observe them long enough. This is called “competing risks”. The researchers created a new way to make predictions about multiple outcomes by taking into account the limited time we have to observe these events. Their method, MultiIncidence, works really well and is much faster than other approaches. It can predict when any of these events might happen in the future. |
Keywords
» Artificial intelligence » Boosting » Optimization
