Active learning based requirement mining for cyber-physical systems

Chen, Gang; Sabato, Zachary; Kong, Zhaodan

doi:10.1109/cdc.2016.7798967

Cited by 15 publications

(14 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lemma 4: For any ρ ∈ (0, 1), if λ t = 2B + 300γ t log 3 (t/ρ), where γ t is the maximum information gain defined in [29], then…”

Section: Theoretical Results Regarding the Effectiveness Of Our Almentioning

confidence: 99%

“…holds with probability > 1 − ρ. Proof: The lemma can be proved by following similar steps as the proof of Lemma 2 in [29].…”

Section: Theoretical Results Regarding the Effectiveness Of Our Almentioning

confidence: 99%

“…It is quite obvious that the quality of the system identification and the system abstraction depends on the quality of the data set. To improve the quality of the system identification (in other words, to decrease the number of data points needed for the system identification), we use an active learning algorithm developed by our group [29] to sample high quality (or "informative") data points for the system identification component after the initial round (see Fig. 1).…”

Section: Active Samplingmentioning

confidence: 99%

See 2 more Smart Citations

Data-Driven Approximate Abstraction for Black-Box Piecewise Affine Systems

Chen

Kong

2018

2018 Annual American Control Conference (ACC)

Self Cite

View full text Add to dashboard Cite

How to effectively and reliably guarantee the correct functioning of safety-critical cyber-physical systems in uncertain conditions is a challenging problem. This paper presents a data-driven algorithm to derive approximate abstractions for piecewise affine systems with unknown dynamics. It advocates a significant shift from the current paradigm of abstraction, which starts from a model with known dynamics. Given a black-box system with unknown dynamics and a linear temporal logic specification, the proposed algorithm is able to obtain an abstraction of the system with an arbitrarily small error and a bounded probability. The algorithm consists of three components, system identification, system abstraction, and active sampling. The effectiveness of the algorithm is demonstrated by a case study with a soft robot.

show abstract

“…Lemma 4: For any ρ ∈ (0, 1), if λ t = 2B + 300γ t log 3 (t/ρ), where γ t is the maximum information gain defined in [29], then…”

Section: Theoretical Results Regarding the Effectiveness Of Our Almentioning

confidence: 99%

“…holds with probability > 1 − ρ. Proof: The lemma can be proved by following similar steps as the proof of Lemma 2 in [29].…”

Section: Theoretical Results Regarding the Effectiveness Of Our Almentioning

confidence: 99%

Section: Active Samplingmentioning

confidence: 99%

See 1 more Smart Citation

Data-Driven Approximate Abstraction for Black-Box Piecewise Affine Systems

Chen

Kong

2018

2018 Annual American Control Conference (ACC)

Self Cite

View full text Add to dashboard Cite

show abstract

“…In this paper we follow the workflow in Algorithm 1, deriving the cost function f ϕ in it from a Bayesian network. For the optimization step (Line 3 of Algorithm 1) we use Gaussian process optimizationwe follow [4,9,10] about this choice. It has a feature that it suggests the global shape of an unknown function; this feature turns out to be convenient for our purpose of integrating causal information in falsification.…”

Section: Gaussian Process Optimizationmentioning

confidence: 99%

Causality-Aided Falsification

Akazaki

Kumazawa

Hasuo

2017

Electron. Proc. Theor. Comput. Sci.

View full text Add to dashboard Cite

Falsification is drawing attention in quality assurance of heterogeneous systems whose complexities are beyond most verification techniques' scalability. In this paper we introduce the idea of causality aid in falsification: by providing a falsification solver-that relies on stochastic optimization of a certain cost function-with suitable causal information expressed by a Bayesian network, search for a falsifying input value can be efficient. Our experiment results show the idea's viability.

show abstract

“…An earlier version of this paper [18] appeared in the 2016 IEEE 55th Conference on Decision and Control (CDC). This paper significantly extends that paper by (i) solving a formal interpretation problem rather than a requirement mining problem, (ii) removing the assumption that the hyper-parameters of the underlying Gaussian process (GP) are known a priori, (iii) providing detailed proofs of theoretical results, and (iv) offering a detailed case study that is closer to real practices.…”

Section: Introductionmentioning

confidence: 99%