Influence of State-Variable Constraints on Partially Observable Monte Carlo Planning
Authors: Alberto Castellini, Georgios Chalkiadakis, Alessandro Farinelli
IJCAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Results on a case study based on Rocksample show that the usage of this knowledge provides significant improvements to the performance of the algorithm. The extent of this improvement depends on the amount of knowledge encoded in the constraints and reaches the 50% of the average discounted return in the most favorable cases that we analyzed. |
| Researcher Affiliation | Academia | 1Computer Science Department, University of Verona, Italy 2School of Electrical and Computer Engineering, Technical University of Crete, Greece |
| Pseudocode | No | The paper does not contain any pseudocode or clearly labeled algorithm blocks. |
| Open Source Code | No | The paper states 'The original C++ code of the POMCP algorithm provided in [Silver and Veness, 2010] was used to perform tests in the standard case (STD). The same code was modified in two ways...', but it does not provide a link or explicit statement about making their modified code open-source. |
| Open Datasets | Yes | We tested our approach on Rocksample(11,11) [Smith and Simmons, 2004], in which 11 rocks are randomly arranged on a grid with 11 rows and 11 columns. |
| Dataset Splits | No | The paper mentions performing '50 runs' and averaging results but does not describe traditional dataset splits (e.g., 80/10/10%) for training, validation, and testing of a fixed dataset. The RockSample problem is a simulation environment. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used to run the experiments. |
| Software Dependencies | No | The paper mentions using 'The original C++ code of the POMCP algorithm' but does not specify version numbers for C++ or any other software dependencies. |
| Experiment Setup | Yes | In particular, we repeated each of the 50 runs for n Sim from 23 = 8 to 214 = 16384, with steps of the power of 2. Test parameters are displayed in Table 1 and corresponding results shown in Figure 2. ... We performed two kinds of experiments on the CSV planner, one to evaluate the influence of the number of state-variable constraints on planning performance, and one to evaluate the influence of uncertainty over state-variable constraints on planning performance. |