Constrained Serial Dictatorships Can Be Fair
Authors: Sylvain Bouveret, Hugo Gilbert, Jérôme Lang, Guillaume Méroué
IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We conclude with experimental results showing how the optimal sequence is impacted by various parameters. (...) 6 Numerical Tests We performed several experiments to explore the properties of the CSDs obtained by maximizing either USW, NSW or ESW. |
| Researcher Affiliation | Academia | 1Univ. Grenoble Alpes, CNRS, Grenoble INP, LIG, 38000 Grenoble, France 2Universit e Paris-Dauphine, Universit e PSL, CNRS, LAMSADE, 75016 Paris, France 3Universit e Cˆote d Azur, Inria, CNRS, I3S, Templier 1, 06410 Biot, France |
| Pseudocode | Yes | Algorithm 1 Greedy ESW Require: the number of agents n, the number of goods m, the scoring vector s, the oracle algorithm TΨ 1: k (0, . . . , 0) # empty CSD 2: max k, max esw k, 0 3: for t = 1 to m do 4: i argmini [n]EU k Ψ(ai) 5: ki ki + 1 6: if SW E Ψ (k) > max esw then 7: max k, max esw k, SW E Ψ (k) 8: end if 9: end for 10: return Completion(max k); |
| Open Source Code | Yes | Code and an interactive demo are available at https://github.com/Guillaume Meroue/CSD-can-be-Fair and https://guillaumemeroue.github.io/IJCAI25. |
| Open Datasets | No | The paper describes methods for generating synthetic data (preference profiles) based on models like Plackett-Luce and Mallows, and then samples from these models for numerical tests. It does not use or provide access to a distinct, pre-existing publicly available dataset in the conventional sense. |
| Dataset Splits | No | The paper generates data by sampling preference profiles from probabilistic models (e.g., 10000 profiles from PLν and Mllφ,µ). It does not explicitly define or use traditional training, validation, or test dataset splits. |
| Hardware Specification | No | The paper does not specify any particular hardware (e.g., CPU, GPU models, memory) used for running the experiments. |
| Software Dependencies | No | The paper mentions the "Pref Sampling library" but does not provide specific version numbers for it or any other software dependencies, which would be necessary for reproducibility. |
| Experiment Setup | Yes | Figure 1 displays the proportion of utility (left-hand side) and goods (right-hand side) obtained for n = 5 and increasing the number of goods m from 5 to 300 in steps of 5. To generate both figures, the IC model and the Borda scoring vector were used and we optimized either USW, ESW or NSW. (...) Figure 2 displays the utility value (plots at the bottom) and the number of items (top) received by each of 5 agents with m = 70 goods, for models PLν (right) and Mllφ,µ (left). In the former model, we use νx = (xm, xm 1, . . . , x1) and decrease x from 1.5 (which already yields very correlated preference profiles similar to FC) to 1 (IC) in steps of 0.01. In the latter model, we increase φ from 0 (FC) to 1 (IC) in steps of 0.02. |