Summary of Combinatorial Logistic Bandits, by Xutong Liu et al.
Combinatorial Logistic Bandits
by Xutong Liu, Xiangxiang Dai, Xuchuang Wang, Mohammad Hajiesmaili, John C.S. Lui
First submitted to arxiv on: 22 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- 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 This paper introduces Combinatorial Logistic Bandits (CLogB), a novel framework for online decision-making. In CLogB, a subset of base arms is selected in each round, with binary outcomes following a logistic parametric model. The framework is applicable to scenarios like online content delivery and dynamic channel allocation. The authors propose two algorithms: CLogUCB, which achieves a regret bound of O(d√(κKT)), and VA-CLogUCB, which attains a regret bound of O(d√KT) under the 1-norm triggering probability modulated (TPM) condition. These results improve upon prior work by factors of √(κ) and √(κ) respectively. The authors also enhance the computational efficiency of VA-CLogUCB by eliminating nonconvex optimization when context features are time-invariant. Experiments on synthetic and real-world datasets demonstrate the superior performance of these algorithms compared to benchmarks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about a new way to make decisions online, called Combinatorial Logistic Bandits (CLogB). Imagine you have many options to choose from, like which movie to watch or which news article to read. CLogB helps you decide by looking at the past and making predictions about what might happen if you choose one option over another. The authors created two new algorithms that use this idea: CLogUCB and VA-CLogUCB. These algorithms are better than others because they make fewer mistakes when choosing options. The authors tested these algorithms on real data and showed that they work well. |
Keywords
* Artificial intelligence * Optimization * Probability
