Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1]
Solving the Torpedo Scheduling Problem
Authors: Martin Josef Geiger, Lucas Kletzander, Nysret Musliu
JAIR 2019 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental investigations are performed on a larger number of problem instances, which stem from the 2016 implementation challenge of the Association of Constraint Programming (ACP). |
| Researcher Affiliation | Academia | Martin Josef Geiger EMAIL Helmut-Schmidt-University University of the Federal Armed Forces Hamburg Holstenhofweg 85, 22043 Hamburg, Germany Lucas Kletzander EMAIL Nysret Musliu EMAIL Christian Doppler Laboratory for Artificial Intelligence and Optimization for Planning and Scheduling DBAI, TU Wien, Karlsplatz 13, 1040 Vienna, Austria |
| Pseudocode | Yes | ALGORITHM 1: Computation of bound ECO; ALGORITHM 2: Computation of bound CCF; ALGORITHM 3: Computation of bound CCMA; ALGORITHM 4: Simulated Annealing; ALGORITHM 5: generate Initial Solution(); ALGORITHM 6: Application of the reduction procedures |
| Open Source Code | No | The paper does not contain a specific statement or link to the open-source code for the methodology described. It refers to conference proceedings but no public code repository. |
| Open Datasets | Yes | An original set of six instances was provided for the ACP 2016 Challenge (Schaus et al., 2016). Those instances were generated by an instance generator that was published along with the configuration files used for the generator. |
| Dataset Splits | No | The paper evaluates its approach on "a larger number of problem instances," specifically mentioning "an original set of six instances" and "additional instances generated by a randomized generator" without defining training, validation, or testing splits within these instances. Each instance is a problem to be solved. |
| Hardware Specification | Yes | The algorithm was executed on an Intel Core i7-6700K with 4 x 4.0 GHz using a single core. |
| Software Dependencies | No | The paper describes its algorithm and methods but does not provide specific version numbers for programming languages, libraries, or solvers used in its implementation. |
| Experiment Setup | Yes | For these results the algorithm was evaluated 20 times on each instance. The evaluation without reduction procedures in Table 5 uses outer = 10000, t fraction = 10 and t factor = 0.998, while the evaluation with reduction procedures in Table 6 uses the much faster cooling with outer = 1000, t fraction = 1000 and t factor = 0.99. |