Additive Merge-and-Shrink Heuristics for Diverse Action Costs

Authors: Gaojian Fan, Martin Müller, Robert Holte

IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our experiments demonstrate that in these domains, an additive set of M&S heuristics using the new cost partitioning method produces much more informative and effective heuristics than creating a single M&S heuristic which directly encodes diverse costs.
Researcher Affiliation Academia Gaojian Fan, Martin M uller and Robert Holte University of Alberta, Canada {gaojian, mmueller, rholte}@ualberta.ca
Pseudocode No The paper describes procedures and mappings in text and tables (e.g., Table 2 showing cost mapping) but does not include structured pseudocode or algorithm blocks.
Open Source Code No The paper does not contain any explicit statements about releasing source code or provide links to a code repository.
Open Datasets No The paper refers to 'nonunit cost IPC domains' and 'GRIPPER' and provides a general competition website ('www.icaps-conference.org/index.php/Main/Competitions') but does not provide a specific link, DOI, repository name, or formal citation with authors/year for dataset access.
Dataset Splits No The paper compares performance on unit-cost and non-unit-cost versions of IPC domain instances, but it does not describe specific training, validation, or test dataset splits (e.g., percentages or counts) for a model.
Hardware Specification No The paper mentions time and memory limits for experiments but does not provide specific details about the hardware (e.g., CPU, GPU models, memory amounts) used.
Software Dependencies No The paper references 'Fast-Downward' (implicitly through its documentation URL) and general algorithms like 'Dijkstra s algorithm' and 'A*', but it does not list specific software dependencies with version numbers (e.g., Python, libraries, or solvers with versions).
Experiment Setup Yes The M&S construction of DCP-MS uses the same configuration as single M&S (the default recommended configuration). The recommended configuration: DFP merging, bisimulation shrinking, label reduction before shrinking, maximum of 50,000 states per abstraction (from www.fast-downward.org/Doc/Heuristic#Merge-and-shrink_heuristic). Each run has a 30 minute time limit and a 2 GB memory limit.