A Robust Genetic Algorithm for Learning Temporal Specifications from Data

Nenzi, Laura; Silvetti, Simone; Bartocci, Ezio; Bortolussi, Luca

doi:10.1007/978-3-319-99154-2_20

Cited by 36 publications

(24 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We do not distinguish these names, sometimes we use them interchangeably, and here we just use the term model learning or learning model instead. Similarly, the learning specification problems also have different names under research, typically are specification mining [38]- [43], specification inference [44], requirements mining [45], mining properties [46], learning logic formulae [47], learning specifications [48], [49], learning properties [50], [51], etc. Here we use the term learning specification (or specification learning) in analogy with learning model (or model learning).…”

Section: Taxonomy Of Learning Algorithms In Formal Methodsmentioning

confidence: 99%

“…But, if neither domain knowledge is available nor the user is not familiar with the system properties that are to be inferred, a step further is to infer the formula structure in addition to its parameters from data. In [58], [59], [48] the first algorithm was proposed to learn both the formula structure and its parameters from data, this approach was called temporal logic inference (TLI).…”

Section: ) With a Model Vs Model Freementioning

confidence: 99%

See 1 more Smart Citation

Survey on Learning-Based Formal Methods: Taxonomy, Applications and Possible Future Directions

et al. 2020

View full text Add to dashboard Cite

Formal methods play an important role in testing and verifying software quality, especially in modern society with rapid technological updates. Learning-based techniques have been extensively applied to learn (a model or model-free) for formal verification and to learn system specifications, and resulted in numerous contributions. Due to the fact that adequate system models are often difficult to design manually and manual definition of specifications for such software systems gets infeasible, which motivate new research directions in learning models and/or specifications from observed system behaviors automatically. This paper mainly concentrates on learning-based techniques in formal methods area. An up-to-date overview of the current state-of-the-art in learning-based formal methods is provided in the paper. This paper is not a comprehensive survey of learning-based techniques in formal methods area, but rather as a survey of the taxonomy, applications and possible future directions in learning-based formal methods.

show abstract

Section: Taxonomy Of Learning Algorithms In Formal Methodsmentioning

confidence: 99%

Section: ) With a Model Vs Model Freementioning

confidence: 99%

Survey on Learning-Based Formal Methods: Taxonomy, Applications and Possible Future Directions

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Proposed inference methods span two categories: parameter synthesis, that is, given the structure of an STL formula to learn parameters to classify signals correctly [9]; and learning both the structure and parameters of the formula. The latter has been approached, e.g., via offline and online supervised learning approaches based on decision trees [10], evolutionary algorithms [11], unsupervised learning approaches [12], or active learning [13], where the authors consider a setting where data is a priori not available but interactively retrieved through queries between the learning algorithm and the signal-producing system.…”

Section: In Robot Motion Planning and Control Desired Behaviormentioning

confidence: 99%

Formalizing Trajectories in Human-Robot Encounters via Probabilistic STL Inference

Linard

Torre

Steen

et al. 2021

2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

View full text Add to dashboard Cite

In this paper, we are interested in formalizing human trajectories in human-robot encounters. We consider a particular case where a human and a robot walk towards each other. A question that arises is whether, when, and how humans will deviate from their trajectory to avoid a collision. These human trajectories can then be used to generate socially acceptable robot trajectories. To model these trajectories, we propose a data-driven algorithm to extract a formal specification expressed in Signal Temporal Logic with probabilistic predicates. We evaluated our method on trajectories collected through an online study where participants had to avoid colliding with a robot in a shared environment. Further, we demonstrate that probabilistic STL is a suitable formalism to depict human behavior, choices and preferences in specific scenarios of social navigation.

show abstract

“…The vast majority of the works in this regard start from a dataset of labeled trajectories, partitioned into positive and negative examples, and try to learn an STL classifier for the data. As an example, Nenzi et al [13] proposed ROGE (RObustness GEnetic algorithm), a bi-level optimization procedure, which optimized the structure by a genetic algorithm and the parameters using Bayesian Optimization. To the best of our knowledge, it is the only attempt at using an evolutionary algorithm for solving the template-free problem.…”

Section: Related Workmentioning

confidence: 99%

Mining Road Traffic Rules with Signal Temporal Logic and Grammar-Based Genetic Programming

2021

Self Cite

View full text Add to dashboard Cite

Traffic systems, where human and autonomous drivers interact, are a very relevant instance of complex systems and produce behaviors that can be regarded as trajectories over time. Their monitoring can be achieved by means of carefully stated properties describing the expected behavior. Such properties can be expressed using Signal Temporal Logic (STL), a specification language for expressing temporal properties in a formal and human-readable way. However, manually authoring these properties is a hard task, since it requires mastering the language and knowing the system to be monitored. Moreover, in practical cases, the expected behavior is not known, but it has instead to be inferred from a set of trajectories obtained by observing the system. Often, those trajectories come devoid of human-assigned labels that can be used as an indication of compliance with expected behavior. As an alternative to manual authoring, automatic mining of STL specifications from unlabeled trajectories would enable the monitoring of autonomous agents without sacrificing human-readability. In this work, we propose a grammar-based evolutionary computation approach for mining the structure and the parameters of an STL specification from a set of unlabeled trajectories. We experimentally assess our approach on a real-world road traffic dataset consisting of thousands of vehicle trajectories. We show that our approach is effective at mining STL specifications that model the system at hand and are interpretable for humans. To the best of our knowledge, this is the first such study on a set of unlabeled real-world road traffic data. Being able to mine interpretable specifications from this kind of data may improve traffic safety, because mined specifications may be helpful for monitoring traffic and planning safety promotion strategies.

show abstract

A Robust Genetic Algorithm for Learning Temporal Specifications from Data

Cited by 36 publications

References 26 publications

Survey on Learning-Based Formal Methods: Taxonomy, Applications and Possible Future Directions

Survey on Learning-Based Formal Methods: Taxonomy, Applications and Possible Future Directions

Formalizing Trajectories in Human-Robot Encounters via Probabilistic STL Inference

Mining Road Traffic Rules with Signal Temporal Logic and Grammar-Based Genetic Programming

Contact Info

Product

Resources

About