Approximate Stream Reasoning with Metric Temporal Logic under Uncertainty

Authors: Daniel de Leng, Fredrik Heintz2760-2767

AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The proposed approach is empirically evaluated by considering the time and space requirements, as well as the impact of permitting varying degrees of uncertainty.
Researcher Affiliation Academia Daniel de Leng, Fredrik Heintz Department of Computer and Information Science Link oping University, 581 83 Link oping, Sweden {daniel.de.leng, fredrik.heintz}@liu.se
Pseudocode Yes Algorithm 1: Classical Progression Algorithm 2: Approximate Partial-State Progression
Open Source Code Yes The jprogress implementation is available at https://github.com/dnleng/jprogress.
Open Datasets No The paper describes generating streams with specific probabilities (e.g., 'stream in which 80% of the samples are p and the remaining samples are unknown'). It does not use a named, publicly available dataset with concrete access information (link, DOI, or formal citation).
Dataset Splits No The paper does not provide specific train/validation/test dataset splits. It describes characteristics of the input stream and termination conditions for the progression procedure.
Hardware Specification Yes We performed our experiments using a fourth-generation Intel Xeon E5-1650 CPU (6 cores, 12 threads) with 50Gi B of RAM allocated to the JVM.
Software Dependencies No The paper states the implementation was 'in Java1' but does not provide a specific version number for Java or any other software dependencies.
Experiment Setup Yes We first compare the runtime and space requirements given a formula and a stream for varying values for the parameters MAX TTL and MAX NODES. ... The choices for MAX NODES limit the leaked mass to at most 1% of the total probability mass.