Learning Stochastic Timed Automata from Sample Executions

Pedro, André de Matos; Crocker, Paul; Sousa, Simão Melo de

doi:10.1007/978-3-642-34026-0_38

Cited by 6 publications

(3 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the last decades, an increasing interest has been shown for learning probabilistic models, partly due to the success of verification techniques such as probabilistic and statistical model checking [5,6]. Despite this development, only few works considered the problem of learning stochastic timed models [11,13,10,9].…”

Section: Related Workmentioning

confidence: 99%

“…In [10], authors focus on learning more general stochastic timed models, namely Generalized Semi-Markov Processes following the same state-merging procedure. This algorithm relies on new statistical clocks estimators, in addition to the state compatibility criterion used in Alergia.…”

Section: Related Workmentioning

confidence: 99%

“…despite the wide development of ML techniques, only few works were interested in learning stochastic timed models [11,13,10,9]. In this paper, we study the RTI+ algorithm [13] and we propose improvements that enhance its accuracy.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Improved Learning for Stochastic Timed Models by State-Merging Algorithms

Mediouni

Nouri

Bozga

et al. 2017

Lecture Notes in Computer Science

View full text Add to dashboard Cite

The construction of faithful system models for quantitative analysis, e.g., performance evaluation, is challenging due to the inherent systems' complexity and unknown operating conditions. To overcome such difficulties, we are interested in the automated construction of system models by learning from actual execution traces. We focus on the timing aspects of systems that are assumed to be of stochastic nature. In this context, we study a state-merging procedure for learning stochastic timed models and we propose several enhancements at the level of the learned model structure and the underlying algorithms. The results obtained on different examples show a significant improvement of timing accuracy of the learned models.

show abstract