Summary of Enhancing Preference-based Linear Bandits Via Human Response Time, by Shen Li et al.
Enhancing Preference-based Linear Bandits via Human Response Time
by Shen Li, Yuyang Zhang, Zhaolin Ren, Claire Liang, Na Li, Julie A. Shah
First submitted to arxiv on: 9 Sep 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Econometrics (econ.EM); 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 In this paper, researchers develop a novel approach to infer human preferences by combining binary choices with response times. They propose a computationally efficient method that leverages the EZ diffusion model from psychology to estimate human utility functions. Theoretical and empirical analyses demonstrate that incorporating response times significantly improves utility estimation for queries with strong preferences. This estimator is then integrated into preference-based linear bandits for fixed-budget best-arm identification. Simulations on three real-world datasets show that using response times accelerates preference learning compared to choice-only approaches. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This study finds a way to better understand what people like and dislike by combining two types of information: how quickly they make choices, and whether they choose one option over another. The researchers use this approach to create a more accurate model of human preferences. They test their method on real-world datasets and find that it works much faster than previous methods. |
Keywords
* Artificial intelligence * Diffusion model
