Sampling Winners in Ranked Choice Voting

Authors: Matthew Iceland, Anson Kahng, Joseph Saber

IJCAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In our experiments, we explore two main questions on a mix of synthetic elections generated from statistical cultures in the map of elections [Szufa et al., 2020; Boehmer et al., 2021] and real-world election data sourced from Pref Lib [Mattei and Walsh, 2013] and the Harvard Dataverse [Harvard, 2020].
Researcher Affiliation Academia Matthew Iceland , Anson Kahng and Joseph Saber University of Rochester miceland@u.rochester.edu, anson.kahng@rochester.edu, jsaber@ur.rochester.edu
Pseudocode No The paper describes methods and processes in paragraph form, but does not include any clearly labeled 'Pseudocode' or 'Algorithm' blocks, figures, or sections.
Open Source Code Yes Our code is available at https://github.com/miceland2/STVsampling.
Open Datasets Yes synthetic elections generated from statistical cultures in the map of elections [Szufa et al., 2020; Boehmer et al., 2021] and real-world election data sourced from Pref Lib [Mattei and Walsh, 2013] and the Harvard Dataverse [Harvard, 2020]
Dataset Splits No The paper describes using various sample sizes (e.g., 5%, 10%, 30%, 50%, 70%, and 100%) and averaging results over multiple samples (e.g., 1,000 samples). However, it does not specify traditional machine learning dataset splits such as explicit train, validation, or test percentages or sample counts for model training and evaluation.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., GPU models, CPU types, memory) used to run the experiments.
Software Dependencies No The paper mentions using the 'mapel Python library' but does not specify its version number or any other software dependencies with their versions.
Experiment Setup Yes In our experiments, we vary the sample size from 10% to 100% in steps of 10%, with the addition of a 5% sample size... The average prediction accuracy is taken over 100 samples on each of 100 different profiles generated according to the statistical cultures in consideration. These profiles each consist of 100 votes over 5 alternatives...