Towards More Practical and Efficient Automatic Dominance Breaking
Authors: Jimmy H.M. Lee, Allen Z. Zhong3868-3876
AAAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experimentation confirms the benefits of the new proposals. Experimental Evaluation We perform experiments on Xeon E7-4830 2.20GHz processors using Mini Zinc 2.4.3 (Nethercote et al. 2007) to model both the benchmark problems and the corresponding dominance breaking nogood generation respectively. The back-end solver is Chuffed (Ohrimenko, Stuckey, and Codish 2009). We use six benchmark problems, and generate 20 instances for each problem configuration in order to give meaningful comparison. |
| Researcher Affiliation | Academia | Jimmy H.M. Lee and Allen Z. Zhong Department of Computer Science and Engineering The Chinese University of Hong Kong, Shatin, N.T., Hong Kong {jlee, zwzhong}@cse.cuhk.edu.hk |
| Pseudocode | No | The paper describes theoretical concepts and proofs but does not include any explicit pseudocode blocks or algorithm listings. |
| Open Source Code | No | The paper states 'The model is from a public Mini Zinc repository1. 1https://people.eng.unimelb.edu.au/pstuckey/dominance/' which refers to a third-party model repository, but it does not provide concrete access to the authors' own source code for the methodology described in this paper. |
| Open Datasets | No | The paper refers to 'six benchmark problems' and states that 'The following four benchmarks and associated models as well as the instance generation method are adopted from Lee and Zhong (2020)'. It also mentions generating '20 instances for each problem configuration'. While it cites a source for the benchmarks, it does not provide concrete access information (e.g., specific links, DOIs, or explicit statements about public availability) for the datasets used in the experiments. |
| Dataset Splits | No | The paper does not provide specific information regarding training/validation/test dataset splits (e.g., percentages, sample counts, or cross-validation setup) for reproducing the data partitioning. |
| Hardware Specification | Yes | We perform experiments on Xeon E7-4830 2.20GHz processors using Mini Zinc 2.4.3 (Nethercote et al. 2007) to model both the benchmark problems and the corresponding dominance breaking nogood generation respectively. |
| Software Dependencies | Yes | We perform experiments on Xeon E7-4830 2.20GHz processors using Mini Zinc 2.4.3 (Nethercote et al. 2007) to model both the benchmark problems and the corresponding dominance breaking nogood generation respectively. The back-end solver is Chuffed (Ohrimenko, Stuckey, and Codish 2009). |
| Experiment Setup | Yes | We use the search strategies specified in the public models in solving the COPs, and the default of Chuffed for the generation models. For all benchmarks, we attempt to generate all dominance breaking nogoods of length up to L without CAE (L-dom) as Lee and Zhong (2020) or with CAE (L-dom(*)) as per our proposal. We use a uniform timeout limit of 1 hour for nogood generation. The timeout for the whole solving process (nogood generation + problem solving) is set to 2 hours, while we keep the timeout for nogood generation as 1 hour. |