An Exact Single-Agent Task Selection Algorithm for the Crowdsourced Logistics
Authors: Chung-Kyun Han, Shih-Fen Cheng
IJCAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In all numerical instances we study, our approach manages to reach optimality yet with much fewer computational resource requirement than the plain integer linear programming formulation. The Numerical Experiments This section explains procedures for generating synthetic problem instances and compares our branch-and-cut algorithm against a greedy heuristic and the ILP model. All approaches are implemented in C++ and tested on identical hardware/software environment (a server with 4 Intel Xeon Gold 6154 CPUs (a total of 72 cores) and 512GB RAM, running Red Hat Linux v6.9). |
| Researcher Affiliation | Academia | Chung-Kyun Han and Shih-Fen Cheng School of Information Systems, Singapore Management University ckhan.2015@phdcs.smu.edu.sg, sfcheng@smu.edu.sg |
| Pseudocode | No | The paper describes algorithmic steps in narrative text (e.g., 'The branch-and-cut (Bn C) approach...', 'The basic idea of our heuristics is...'), but it does not include formal pseudocode blocks or figures explicitly labeled as 'Pseudocode' or 'Algorithm'. |
| Open Source Code | No | The paper does not provide any statement about releasing source code or a link to a code repository for the described methodology. |
| Open Datasets | No | The paper explicitly states that it generates 'random problem instances based on the following parameters', implying synthetic data rather than a publicly available dataset. No link or citation to an existing public dataset is provided. |
| Dataset Splits | No | The paper generates 'random numerical instances' for its experiments and does not describe explicit training, validation, or test dataset splits in the conventional sense of partitioning a single dataset. |
| Hardware Specification | Yes | All approaches are implemented in C++ and tested on identical hardware/software environment (a server with 4 Intel Xeon Gold 6154 CPUs (a total of 72 cores) and 512GB RAM, running Red Hat Linux v6.9). |
| Software Dependencies | Yes | We use CPLEX 12.8, a commercial solver, to solve ILP models using default settings. |
| Experiment Setup | Yes | We generate random problem instances based on the following parameters: nr: The number of routine nodes, nr = |R|. np: The number of pickup nodes, np = |P|. nd: The number of delivery nodes, nd = |D|. ca: The agent s volume and weight capacity limits. dt: The additional detour time the agent is willing to spend on delivery tasks, expressed as the percentage of traveling time to complete the agent s routine route. For all our numerical experiments, we fix nr = 6, np = 4, and dt = 25%, and vary nd and ca. For each configuration (nd, ca), we generate 20 random numerical instances. |