Summary of Evaluating and Correcting Performative Effects Of Decision Support Systems Via Causal Domain Shift, by Philip Boeken et al.
Evaluating and Correcting Performative Effects of Decision Support Systems via Causal Domain Shift
by Philip Boeken, Onno Zoeter, Joris M. Mooij
First submitted to arxiv on: 1 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Statistics Theory (math.ST)
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 This paper tackles performative predictions in algorithmic decision support systems (DSS), where an agent’s actions affect the outcome. The authors highlight the importance of assessing these effects, especially in high-stakes settings like healthcare or predictive policing. They propose modeling DSS deployment as causal domain shift and develop novel methods for estimating conditional expectations, allowing for pre- and post-hoc assessment of the system’s impact. A repeated regression procedure is shown to be effective in handling sample selection bias and selective labeling. The paper presents a practical framework for addressing multiple forms of target variable bias. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary In this research, scientists are trying to improve how computers make predictions that affect real-world decisions. They want to make sure these predictions don’t accidentally change the outcome they’re supposed to predict. This is important because predictions can influence things like who gets medical treatment or whether someone gets arrested. The team proposes a new way to analyze how predictions work when they’re actually used in the world, and shows that their method can handle some common problems that make it harder to get accurate results. |
Keywords
* Artificial intelligence * Regression
