Summary of Queueing Matching Bandits with Preference Feedback, by Jung-hun Kim et al.
Queueing Matching Bandits with Preference Feedback
by Jung-hun Kim, Min-hwan Oh
First submitted to arxiv on: 14 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 Medium Difficulty summary: This study investigates multi-class multi-server asymmetric queueing systems, where jobs arrive at each time in unknown queues and are assigned to servers based on their preferences. The service rates of servers are modeled using a feature-based Multi-nomial Logit (MNL) function, and the goal is to stabilize the queues while learning these rates. To achieve this, the authors propose algorithms based on Upper Confidence Bound (UCB) and Thompson Sampling, which provide system stability with an average queue length bound of O(min{N,K}/ε) for a large time horizon T, where ε is the traffic slackness. The algorithms also achieve sublinear regret bounds of Õ(min{√T Q_max,T^(3/4)}), where Q_max represents the maximum queue length over agents and times. Experimental results demonstrate the performance of these algorithms. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Low Difficulty summary: This study looks at a way to manage jobs that come in to be processed by multiple servers, where each job has different needs. The researchers develop special computer programs, called algorithms, to make sure the jobs are handled efficiently and fairly. They test their algorithms on fake data and show they work well. The goal is to make sure the system runs smoothly and doesn’t get too backed up. |
