Dynamic Control of Probabilistic Simple Temporal Networks

Authors: Michael Gao, Lindsay Popowski, Jim Boerkoel9851-9858

AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We empirically compare our approaches against two existing PSTN offline dispatch approaches and one online approach and show that our MIN-LOSS DC algorithm outperforms the others in terms of maximizing execution success while maintaining competitive runtimes.
Researcher Affiliation Academia Michael Gao, Lindsay Popowski, James C. Boerkoel Jr. Human Experience and Agent Teamwork Lab (heatlab.org) Harvey Mudd College Claremont, California 91711 {mgao, lpopowski, boerkoel}@hmc.edu
Pseudocode Yes Algorithm 1: Min-Loss DC; Algorithm 2: MAX-GAIN DC Search
Open Source Code Yes All code and problem instances are available at http://github.com/HEATlab/DCfor PSTN.
Open Datasets Yes We evaluate our methods on two PSTN benchmarks. The first is the DREAM benchmark of 540 PSTNs (Abrahams et al. 2019)... The second is the adapted version of the CAR-SHARING dataset originally from Santana et al. (2016), which was edited by Akmal et al. (2019)...
Dataset Splits No The paper mentions evaluating methods on benchmarks and simulating dispatch 200 times per problem instance, but does not specify explicit training, validation, or testing dataset splits, nor does it describe cross-validation setups.
Hardware Specification No The paper does not explicitly describe the specific hardware used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details with version numbers, such as library or solver names and their versions.
Experiment Setup Yes Interestingly, we found that MIN-LOSS performed best when set to the lowest risk level we tested, α = 0.001, which corresponded to extracting an STNU where the bounds over uncertain edges each capture 99.9% of the likelihood.