A short note about the learning-augmented secretary problem

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Choo, Davin, Ling, Chun Kai
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912101975457792
author Choo, Davin
Ling, Chun Kai
author_facet Choo, Davin
Ling, Chun Kai
contents We consider the secretary problem through the lens of learning-augmented algorithms. As it is known that the best possible expected competitive ratio is $1/e$ in the classic setting without predictions, a natural goal is to design algorithms that are 1-consistent and $1/e$-robust. Unfortunately, [FY24] provided hardness constructions showing that such a goal is not attainable when the candidates' true values are allowed to scale with $n$. Here, we provide a simple and explicit alternative hardness construction showing that such a goal is not achievable even when the candidates' true values are constants that do not scale with $n$.
format Preprint
id arxiv_https___arxiv_org_abs_2410_06583
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A short note about the learning-augmented secretary problem
Choo, Davin
Ling, Chun Kai
Data Structures and Algorithms
We consider the secretary problem through the lens of learning-augmented algorithms. As it is known that the best possible expected competitive ratio is $1/e$ in the classic setting without predictions, a natural goal is to design algorithms that are 1-consistent and $1/e$-robust. Unfortunately, [FY24] provided hardness constructions showing that such a goal is not attainable when the candidates' true values are allowed to scale with $n$. Here, we provide a simple and explicit alternative hardness construction showing that such a goal is not achievable even when the candidates' true values are constants that do not scale with $n$.
title A short note about the learning-augmented secretary problem
topic Data Structures and Algorithms
url https://arxiv.org/abs/2410.06583