Heuristics for Numeric Planning via Subgoaling

Authors: Enrico Scala, Patrik Haslum, Sylvie Thiébaux

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

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments show the effectiveness of its inadmissible and admissible version on satisficing and optimal numeric planning, respectively.
Researcher Affiliation Collaboration 1Research School of Computer Science, The Australian National University 2NICTA Canberra, ACT, Australia firstname.lastname@anu.edu.au
Pseudocode No The paper does not contain any explicitly labeled pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any explicit statements about releasing source code or links to a code repository for the methodology described.
Open Datasets Yes Benchmarks include IPC domains (http://www.icaps-conference.org/), numeric reformulation of the COUNTERS and GARDENING domains by Franc es and Geffner (2015), our motivating example (SAILING) and a new domain called FARMLAND.
Dataset Splits No The paper discusses evaluating heuristics on various planning domains and instances (e.g., SAILING instances scaled by number of boats and people, GARDENING with up to 3 plants), but it does not specify explicit training, validation, and test dataset splits with percentages or sample counts for reproducibility in the context of model training.
Hardware Specification No The paper does not provide specific details regarding the hardware (e.g., CPU, GPU models, memory) used for running the experiments.
Software Dependencies No The paper mentions using concepts similar to Metric-FF and PDDL 2.1 but does not list specific software dependencies with version numbers for its implementation or experimental environment.
Experiment Setup Yes We evaluate three planners: two satisficing planners, using the inadmissible additive heuristic ˆhaddhbd+ and its extension with redundant constraints ˆhraddhbd+, and one optimal planner using the admissible heuristic with redundant constraints ˆhmaxhbd+. The latter is used within A?, while all other heuristics are used in a Greedy Best First Search (GBFS). For comparison, we also run GBFS with the interval-based relaxation (IBR) heuristic, obtained from Metric-FF (Hoffmann 2003) by disabling Enforced Hill Climbing. The optimal planner is compared with blind search. The timeout for all planners was 1, 800 seconds.