A Multiagent Path Search Algorithm for Large-Scale Coalition Structure Generation
Authors: Redha Taguelmimt, Samir Aknine, Djamila Boukredera, Narayan Changder, Tuomas Sandholm
AAAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We show empirically on nine standard value distributions, including disaster response and electric vehicle allocation benchmarks, that our algorithm enables a rapid finding of high-quality solutions and compares favorably with other state-of-the-art methods. |
| Researcher Affiliation | Collaboration | 1Clermont Auvergne University, Clermont Auvergne INP, CNRS, LIMOS, F-63000 Clermont-Ferrand, France. 2Univ Lyon, UCBL, CNRS, INSA Lyon, Centrale Lyon, Univ Lyon 2, LIRIS, UMR5205, Lyon, France 3Laboratory of Applied Mathematics, Faculty of Exact Sciences, University of Bejaia, Bejaia, Algeria 4TCG Centres for Research and Education in Science and Technology, Kolkata, India 5Carnegie Mellon University, Computer Science Department, Pittsburgh, USA 6Strategy Robot, Inc. 7Strategic Machine, Inc. 8Optimized Markets, Inc. |
| Pseudocode | Yes | Algorithm 1 shows the pseudocode of SALDAE. |
| Open Source Code | No | The paper mentions that they used 'the codes provided by the authors of PICS and CSG-UCT', which refers to third-party code, not their own for SALDAE. There is no explicit statement or link indicating that the source code for SALDAE is publicly available. |
| Open Datasets | No | The paper refers to 'nine standard value distributions' (e.g., Agent-based Uniform (Rahwan, Michalak, and Jennings 2012), Beta, Exponential, Gamma (Michalak et al. 2016)) and a 'Disaster Response distribution introduced in (Wu and Ramchurn 2020)' for generating problem instances. However, it does not provide concrete access information (link, DOI, repository) for a specific dataset used, but rather describes distributions for creating problem instances. |
| Dataset Splits | No | The paper states, 'The result for each value distribution was produced by computing the average result from 50 generated problem instances per value distribution.' It does not provide specific training/test/validation dataset splits, percentages, or counts, as it generates problem instances rather than using pre-split datasets. |
| Hardware Specification | Yes | The algorithms were run on an Intel Xeon 2.30GHz E5-2650 CPU with 256GB of RAM. |
| Software Dependencies | No | The paper states, 'We implemented our algorithm in Java' and 'the codes provided by the authors of PICS and CSG-UCT, which are also written in Java.' However, no specific version numbers for Java or any other software libraries/dependencies are provided. |
| Experiment Setup | Yes | In these experiments, we set the number of search agents of SALDAE to 10 and we stopped the algorithms when they finish or at the time when ODP-IP finds the optimal solution in case they take more time to finish than ODP-IP. Moreover, the number of cores used for each algorithm was matched to the number of processes it utilized. Specifically, SALDAE was run on 10 cores, PICS on 20 cores, and CSG-UCT on 1 core. |