Open Loop Execution of Tree-Search Algorithms
Authors: Erwan Lecarpentier, Guillaume Infantes, Charles Lesire, Emmanuel Rachelson
IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We empirically demonstrate that our method achieves a compromise between loss of performance and computational gain. We compared OLUCT with OLTA on a discrete 1D track environment2 and a continuous Physical Travelling Salesman Problem3 (PTSP) [Perez et al., 2012b]. |
| Researcher Affiliation | Collaboration | 1 ONERA The French Aerospace Lab, Toulouse, France 2 ISAE SUPAERO, University of Toulouse, France |
| Pseudocode | Yes | Algorithm 1: OLUCT tree building procedure Algorithm 2: OLTA algorithm |
| Open Source Code | Yes | Code available at: 2https://github.com/erwanlecarpentier/1dtrack.git 3https://github.com/erwanlecarpentier/flatland.git |
| Open Datasets | Yes | Code available at: 2https://github.com/erwanlecarpentier/1dtrack.git 3https://github.com/erwanlecarpentier/flatland.git and [Perez et al., 2012b] Diego Perez, Philipp Rohlfshagen, and Simon M. Lucas. The physical travelling salesman problem: WCCI 2012 competition. |
| Dataset Splits | No | The paper describes simulation settings and generation of episodes for evaluation, but does not provide explicit training, validation, and test dataset splits in the traditional sense, as it applies to planning algorithms that generate data on-the-fly rather than using a fixed dataset. |
| Hardware Specification | No | No specific hardware details (such as GPU or CPU models, memory, or cloud instance types) used for running experiments are provided in the paper. |
| Software Dependencies | No | No specific software dependencies with version numbers (e.g., Python 3.8, PyTorch 1.9) are provided in the paper. |
| Experiment Setup | Yes | The simulation settings are: q {0.0, 0.05, , 0.5}; n = 20 (budget); πdefault = πoptimal; H = 10 (simulation horizon for πdefault); Cp = 0.7; γ = 0.9. The decision criteria parameters were tuned to: τSDM = 80; τSDV = 0.4; τSDSD = 1; τRDV = 0.9. (for 1D Track) and The simulation settings are: s0 = (1.1, 1.1, 0, 0.1); q {0.0, 0.05, , 0.5}; σnoise = 0.02; n = 300 (initial tree budget); πdefault = πgo straight that applies no orientation variation; H = 50 (simulation horizon for πdefault); Cp = 0.7; γ = 0.99. The different decision criteria parameters were tuned to: τSDV = 0.02; τSDSD = 1; τRDV = 0.1. (for PTSP) |