Group Decision Diagram (GDD): A Compact Representation for Permutations

Authors: Takanori Maehara, Yuma Inoue2986-2994

AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conducted experiments to establish that the GDD is more efficient in representing a set of permutations, compared to the πDD and ρDD (Section 5).
Researcher Affiliation Collaboration 1RIKEN AIP, 2Google Japan
Pseudocode Yes Algorithm 1 Naive Schreier Sims algorithm. Algorithm 2 Create a node for (i, α, lo, hi) Algorithm 3 Create a GDD for a singleton. Algorithm 4 Membership predicate. Algorithm 5 Union of two sets. Algorithm 6 Right multiplication. Algorithm 7 Cartesian product.
Open Source Code Yes Our code is available at the project page2. 2https://github.com/spaghetti-source/gdd
Open Datasets No The paper describes using specific problem instances (Rubik's Cube, Pancake Sorting, Shortest Zero-Walk on generated graphs) but does not refer to them as publicly available datasets with specific access information like URLs or DOIs.
Dataset Splits No The paper describes experiments for different problems but does not mention specific train/validation/test splits, as the nature of the problems (e.g., permutation puzzles, group theory) does not involve traditional dataset splitting for model training/evaluation.
Hardware Specification Yes All the experiments were performed on a machine with Intel(R) Xeon(R) CPU E5-2690 v4, 2.60GHz with a single core, with 50GB of RAM.
Software Dependencies Yes The codes was implemented in C++. For comparison, we obtained the original source codes of πDD and ρDD from the authors of the papers (Minato 2011; Inoue and Minato 2014), which are also implemented in C++. These were compiled by g++ (GCC) 4.8.5 with the -O3 option.
Experiment Setup Yes We conducted experiments to evaluate the performance of the GDD. The codes was implemented in C++. For comparison, we obtained the original source codes of πDD and ρDD from the authors of the papers (Minato 2011; Inoue and Minato 2014), which are also implemented in C++. These were compiled by g++ (GCC) 4.8.5 with the -O3 option. All the experiments were performed on a machine with Intel(R) Xeon(R) CPU E5-2690 v4, 2.60GHz with a single core, with 50GB of RAM. Our code is available at the project page2. Also, some additional experimental results are available in the project page, which are omitted from this manuscript due to the space limitation. Rubik’s Cube (Pocket Cube) Puzzle... Pancake Sorting Problem... Shortest Zero-Walk in Group-Labeled Graphs.