Chance-Constrained Probabilistic Simple Temporal Problems

Authors: Cheng Fang, Peng Yu, Brian Williams

AAAI 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this paper we present the probabilistic Simple Temporal Network (p STN), a probabilistic formalism for representing temporal problems with bounded risk and a utility over event timing. We introduce a constrained optimisation algorithm for p STNs that achieves compactness and efficiency through a problem encoding in terms of a parameterised STNU and its reformulation as a parameterised STN. We demonstrate through a car sharing application that our chance-constrained approach runs in the same time as the previous probabilistic approach, yields solutions with utility improvements of at least 5% over previous arts, while guaranteeing operation within the specified risk bound.
Researcher Affiliation Academia Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory 32 Vassar Street, Cambridge, MA 02139 {cfang,yupeng,williams}@mit.edu
Pseudocode Yes Algorithm 1: Approximating cc-p STP
Open Source Code No The paper does not provide any links to open-source code or explicit statements about releasing code.
Open Datasets No The paper states:
Dataset Splits No The paper mentions generating 1800 p STNs but does not specify training, validation, or test splits.
Hardware Specification No The paper mentions
Software Dependencies No The paper mentions
Experiment Setup Yes In each scenario we schedule for a 6 hour period, with the number of cars ranging from 1 to 20, each with up to 5 users. For each user, up to three goal locations were generated based on a simplified open source map of Boston. A p STN was generated for each scenario. The traversal activities were modelled as normally distributed uncertain durations, with the means of u Dns determined by length and speed limits of the roads taken, and standard deviations at 5% of the mean. A total of 1800 p STNs were generated. ... For each p STN, we constructed three cc-p STPs, with chance-constraints 10%, 20% and 40%.