Equitable Stable Matchings in Quadratic Time

Authors: Nikolaos Tziavelis, Ioannis Giannakopoulos, Katerina Doka, Nectarios Koziris, Panagiotis Karras

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our experimental study with simulated markets shows that DLS outperforms the state of the art in equity measures and matches the most efficient ones in runtime.
Researcher Affiliation Academia Nikolaos Tziavelis Northeastern University Ioannis Giannakopoulos NTU Athens Katerina Doka NTU Athens Nectarios Koziris NTU Athens Panagiotis Karras Aarhus University
Pseudocode Yes Algorithm 1 Power Balance
Open Source Code Yes Code and data are available at https://github.com/ntzia/stable-marriage
Open Datasets Yes We use synthetic datasets that draw preferences from three distributions: Uniform(U), Discrete(D), and Gaussian(G)... Last, we apply our solution on real data... Code and data are available at https://github.com/ntzia/stable-marriage
Dataset Splits No The paper mentions using 'synthetic datasets' and 'real data' but does not specify explicit train/validation/test splits, percentages, or sample counts for these datasets.
Hardware Specification Yes The algorithms are implemented in Java4 and tested on an Intel Xeon 2.67GHz CPU with 28GB RAM.
Software Dependencies No The paper states 'The algorithms are implemented in Java4' but does not provide specific Java versions or other software dependencies with version numbers.
Experiment Setup Yes We set the probability parameter in BILS to 0, in i BILS to 0.125, the limit parameter of POWERBALANCE to n log2 2 n/10 , and the parameters in HMS to k = 2 log n and m = log n .