2022•arXiv (Cornell University)Open access

Two 6-approximation Algorithms for the Stochastic Score Classification Problem

Naifeng Liu

Open full text 0 citations

Abstract

We study the arbitrary cost case of the unweighted Stochastic Score Classification (SSClass) problem. We show two constant approximation algorithms and both algorithms are 6-approximation non-adaptive algorithms with respect to the optimal adaptive algorithm. The first algorithm uses a modified round-robin approach among three sequences, which is inspired by a recent result on the unit cost case of the SSClass problem. The second algorithm is originally from the work of Gkenosis et al. In our work, we successfully improve its approximation factor from 2(B-1) to 6. Our analysis heavily uses the relation between computation and verification of functions, which was studied in the information theory literature.

Open-access reader

About this research paper

What this paper is about

We study the arbitrary cost case of the unweighted Stochastic Score Classification (SSClass) problem. We show two constant approximation algorithms and both algorithms are 6-approximation non-adaptive algorithms with respect to the optimal adaptive algorithm. The first algorithm uses a modified round-robin approach among three sequences, which is inspired by a recent result on the unit cost case of the SSClass problem. The second algorithm is originally from the work of Gkenosis et al. In our work, we successfully improve its approximation factor from 2(B-1) to 6. Our analysis heavily uses the relation between computation and verification of functions, which was studied in the information theory literature.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

We study the arbitrary cost case of the unweighted Stochastic Score Classification (SSClass) problem. We show two constant approximation algorithms and both algorithms are 6-approximation non-adaptive algorithms with respect to the optimal adaptive algorithm. The first algorithm uses a modified round-robin approach among three sequences, which is inspired by a recent result on the unit cost case of the SSClass problem. The second algorithm is originally from the work of Gkenosis et al. In our work, we successfully improve its approximation factor from 2(B-1) to 6. Our analysis heavily uses the relation between computation and verification of functions, which was studied in the information theory literature.

Key concepts: Approximation algorithm, Stochastic approximation, Computation, Algorithm, Computer science, Relation (database), Constant (computer programming), Mathematical optimization

Related papers

Back to paper searchBrowse research topicsOriginal source
Two 6-approximation Algorithms for the Stochastic Score Classification Problem — Research Paper | ScholarLens