Individual Fairness in Graph Decomposition

Authors: Kamesh Munagala, Govind S. Sankar

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

Reproducibility Variable Result LLM Response
Research Type Experimental We perform experiments on real-world planar graphs modeling precinct connectivity in redistricting applications in the states of North Carolina, Maryland, and Pennsylvania in the United States.
Researcher Affiliation Academia 1Computer Science Department, Duke University, Durham NC 27708-0129, USA. Correspondence to: Kamesh Munagala <kamesh@cs.duke.edu>, Govind S. Sankar <govind.subash.sankar@duke.edu>.
Pseudocode Yes The paper includes several clearly labeled algorithm blocks: 'Algorithm 1 KPR Decomposition (Klein et al., 1993)', 'Algorithm 2 KPR algorithm with random vertex weights', 'Algorithm 3 KPR algorithm with two cuts per phase', 'Algorithm 4 KPR algorithm with random diameter', and 'Algorithm 5 LDD via Euclidean Embedding'.
Open Source Code No The paper does not contain an explicit statement or a link indicating that the source code for the described methodology is publicly available. While it mentions the MGGG Lab GitHub for data, it does not offer code for their algorithms.
Open Datasets Yes We perform experiments on real-world planar graphs modeling precinct connectivity in redistricting applications in the states of North Carolina, Maryland, and Pennsylvania in the United States. [...] MGGG Lab. Mggg state shapefiles. https://github.com/mggg-states. Accessed January 2024.
Dataset Splits No The paper does not explicitly provide training/test/validation dataset splits (e.g., percentages, sample counts, or references to predefined splits) for reproducing the experiments. The experiments are conducted on entire graph datasets rather than split portions.
Hardware Specification No The paper states 'We implemented Algorithm 2' and 'We ran experiments' but does not provide any specific details about the hardware used, such as GPU/CPU models, memory, or cloud resources.
Software Dependencies No The paper does not list specific software dependencies with version numbers (e.g., 'Python 3.8', 'PyTorch 1.9', 'CPLEX 12.4') that would be necessary to replicate the experiments.
Experiment Setup Yes We implemented Algorithm 2 with parameter R = 5 and compared it with Algorithm 1 with the same R. We plot the histograms of the edge separation probabilities in Figure 2, showing clearly that Algorithm 2 is more individually fair than Algorithm 1. [...] We plot the number of clusters for 200 runs in Figure 3, and note that the empirically observed value is consistently smaller than n/R. [...] In our experiments, we choose the roots for the BFS trees uniformly at random.