Fair and Useful Cohort Selection

Authors: Konstantina Bairaktari, Paul Tsela Langton, Huy Nguyen, Niklas Smedemark-Margulies, Jonathan Ullman

TMLR 2023 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical Our technical contributions are polynomial-time algorithms for fair and useful cohort selection with both linear and ratio utilities, in two different algorithmic models. Offline Setting. In this setting, we are given all the scores s1, . . . , sn at once, and must choose the optimal cohort. We present a polynomial-time algorithm for each model that computes a fair cohort with optimal expected utility. Online Setting. In this setting, we are given the scores s1, s2, . . . as a stream... For ratio utilities we give a fair and optimal algorithm in this setting... For linear utilities, we give an approximately fair and optimal algorithm... We start with a brief overview of the key steps in our algorithms for each setting. The formal description of the algorithms can be found in Sections 3 and 4. All omitted proofs are in the appendix.
Researcher Affiliation Academia Konstantina Bairaktari EMAIL Khoury College of Computer Sciences Northeastern University; Paul Langton EMAIL Khoury College of Computer Sciences Northeastern University; Huy Le Nguyen EMAIL Khoury College of Computer Sciences Northeastern University; Niklas Smedemark-Margulies EMAIL Khoury College of Computer Sciences Northeastern University; Jonathan Ullman EMAIL Khoury College of Computer Sciences Northeastern University
Pseudocode Yes Alg. 1: Offline Cohort Selection for Linear Utility; Alg. 2: Online Cohort Selection for Linear Utility; Alg. 3: Offline Cohort Selection for Ratio Utility; Alg. 4: Online Cohort Selection for Ratio Utility; Algorithm 5: Rounding
Open Source Code No The paper does not provide any explicit statement about releasing source code or links to a code repository for the described methodology.
Open Datasets No The paper focuses on theoretical algorithms for 'fair cohort selection' using 'scores si' for 'n possible candidates'. It does not mention the use of any specific named public datasets, nor does it provide any links or citations to datasets for experimental evaluation.
Dataset Splits No The paper does not describe any specific datasets used for empirical evaluation, and therefore does not mention any dataset splits (e.g., training/test/validation).
Hardware Specification No The paper is theoretical, focusing on algorithm design and proofs. It does not contain any mention of specific hardware (like GPU/CPU models, memory, or cloud resources) used for running experiments.
Software Dependencies No The paper is theoretical, focusing on algorithm design and proofs. It does not mention any specific software dependencies or their version numbers that would be required to replicate experiments or implementations.
Experiment Setup No The paper describes theoretical algorithms and proofs, without presenting empirical experiments. Consequently, there are no details provided regarding an experimental setup, hyperparameters, or system-level training settings.