External Memory Bidirectional Search

Authors: Nathan R. Sturtevant, Jingwei Chen

IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results show the effectiveness of the approach.
Researcher Affiliation Academia Nathan R. Sturtevant Department of Computer Science University of Denver Denver, USA sturtevant@cs.du.edu Jingwei Chen Department of Computer Science University of Denver Denver, USA jingwei.chen@du.edu
Pseudocode Yes Algorithm 1 Generic Best-First Search, Algorithm 2 Generic Bidirectional Search, Algorithm 3 Parallel External-Memory MM, Algorithm 4 Parallel Bucket expansion pseudo-code
Open Source Code No The paper does not provide a direct statement or link indicating that the source code for the described methodology is publicly available.
Open Datasets Yes Our experimental results are in the domain of Rubik s Cube. In addition to solving the standard 10 Korf instances [Korf, 1997], we also solve the superflip position, one of the known problems at depth 20.
Dataset Splits No The paper evaluates performance on specific Rubik's Cube problem instances and does not describe dataset splits (e.g., train/validation/test percentages or sample counts) for reproducibility.
Hardware Specification Yes Our experiments are run on a 2.4 GHz Intel Xeon E5 with dual 8-core processors. The machine has 128 GB of RAM and has two 8TB disk drives on which the experiments are performed.
Software Dependencies No The paper describes the algorithms and their implementation details but does not provide specific version numbers for any software dependencies or libraries used.
Experiment Setup Yes The brute-force results all use 128 buckets, except for the superflip position which used 512. The heuristic search had a baseline of 128 buckets that were further divided by heuristic values, so the exact number of buckets varied on each problem instance.