Research map / Learning theory and algorithms / Combinatorial and search optimization
A*+BFHS: A Hybrid Heuristic Search Algorithm
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Replace BFIDA* with A*+BFHS, using an A* frontier followed by multiple BFHS calls ordered deepest-first | heuristic search on hard unit-cost planning instances · 32 hardest IPC planning instances | running time: 27 of those 32 instances | Helped |
| Replace BFIDA* with A*+BFHS, using an A* frontier followed by BFHS | heuristic search on hard unit-cost planning instances · 32 hardest IPC planning instances | Peak Stored Nodes: less memory than BFIDA* on 25 of the 32 instances | Helped |
| Use A*+BFHS instead of A* when memory is constrained | heuristic search on hard unit-cost planning instances · 32 hardest IPC planning instances | running time: slower than A* | Hurt |
| Use A*+BFHS instead of A* when memory is constrained | heuristic search on hard unit-cost planning instances · 32 hardest IPC planning instances | Peak Stored Nodes: more than 8 GB of memory on one instance | Helped |
| Group frontier nodes at multiple depths in each BFHS call, with at most four calls per iteration (ablation) | A*+BFHS · 32 hardest IPC planning instances | running time: faster by a factor of 3 on logistics00 15-1; 2.5 on rovers 09 and 11; 4.6 on termes18 05; 3.9 on tpp 11 | Helped |
| Group frontier nodes at multiple depths instead of calling BFHS on a single depth (ablation) | A*+BFHS · 32 hardest IPC planning instances | duplicate nodes generated in previous iterations: at least twice as many nodes on 7 instances | Helped |
| Group frontier nodes at multiple depths instead of calling BFHS on a single depth (ablation) | A*+BFHS · 32 hardest IPC planning instances | Peak Stored Nodes: A*+BFHS (4) stored a larger maximum number of nodes on almost all instances | Hurt |
| Call BFHS on frontier nodes at a single depth rather than grouping adjacent depths (ablation) | A*+BFHS · 32 hardest IPC planning instances | nodes generated in the last iteration: A*+BFHS (∞) generated significantly fewer nodes in the last iteration ... on 22 instances | Helped |
| Call BFHS on frontier nodes at a single depth rather than grouping adjacent depths (ablation) | A*+BFHS · 32 hardest IPC planning instances | running time: A*+BFHS (4) was faster by a factor of 3 on logistics00 15-1, 2.5 on rovers 09 and 11, 4.6 on termes18 05, 3.9 on tpp 11 | Mixed |
| Allow only one BFHS call per iteration (ablation) | A*+BFHS · unit-cost search | early termination | Hurt |
Closest papers by meaning
- PEA*+IDA*: An Improved Hybrid Memory-Restricted Algorithm (AAAI 2022 · similarity 0.87)
- New Results in Bounded-Suboptimal Search (AAAI 2022 · similarity 0.85)
- Multi-Objective Evolution of Heuristic Using Large Language Model (AAAI 2025 · similarity 0.84)
- Eco Search: A No-delay Best-First Search Algorithm for Program Synthesis (AAAI 2025 · similarity 0.84)
- Anchor Search: A Unified Framework for Suboptimal Bidirectional Search (AAAI 2025 · similarity 0.83)
- Thought of Search: Planning with Language Models Through The Lens of Efficiency (NeurIPS 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.