TY - GEN
T1 - Safe inputs approximation for black-box systems
AU - Xue, Bai
AU - Liu, Yang
AU - Ma, Lei
AU - Zhang, Xiyue
AU - Sun, Meng
AU - Xie, Xiaofei
N1 - Funding Information:
Acknowledgements. Bai Xue was funded by CAS Pioneer Hundred Talents Program under grant No. Y8YC235015, NSFC under grant No. 61872341 and 61836005, Lei Ma was supported by JSPS KAKENHI Grant 19H04086 and Qdai-jump Research Program NO.01277, Meng Sun was supported by NSFC under grant No. 61772038, 61532019 and 61272160 and the Guangdong Science and Technology Department (Grant no. 2018B010107004).
Publisher Copyright:
© 2019 IEEE.
PY - 2019/11
Y1 - 2019/11
N2 - Given a family of independent and identically distributed samples extracted from the input region and their corresponding outputs, in this paper we propose a method to under-approximate the set of safe inputs that lead the black-box system to respect a given safety specification. Our method falls within the framework of probably approximately correct (PAC) learning. The computed under-approximation comes with statistical soundness provided by the underlying PAC learning process. Such a set, which we call a PAC under-approximation, is obtained by computing a PAC model of the black-box system with respect to the specified safety specification. In our method, the PAC model is computed based on the scenario approach, which encodes as a linear program. The linear program is constructed based on the given family of input samples and their corresponding outputs. The size of the linear program does not depend on the dimensions of the state space of the black-box system, thus providing scalability. Moreover, the linear program does not depend on the internal mechanism of the black-box system, thus being applicable to systems that existing methods are not capable of dealing with. Some case studies demonstrate these properties, general performance and usefulness of our approach.
AB - Given a family of independent and identically distributed samples extracted from the input region and their corresponding outputs, in this paper we propose a method to under-approximate the set of safe inputs that lead the black-box system to respect a given safety specification. Our method falls within the framework of probably approximately correct (PAC) learning. The computed under-approximation comes with statistical soundness provided by the underlying PAC learning process. Such a set, which we call a PAC under-approximation, is obtained by computing a PAC model of the black-box system with respect to the specified safety specification. In our method, the PAC model is computed based on the scenario approach, which encodes as a linear program. The linear program is constructed based on the given family of input samples and their corresponding outputs. The size of the linear program does not depend on the dimensions of the state space of the black-box system, thus providing scalability. Moreover, the linear program does not depend on the internal mechanism of the black-box system, thus being applicable to systems that existing methods are not capable of dealing with. Some case studies demonstrate these properties, general performance and usefulness of our approach.
UR - http://www.scopus.com/inward/record.url?scp=85074656794&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85074656794&partnerID=8YFLogxK
U2 - 10.1109/ICECCS.2019.00027
DO - 10.1109/ICECCS.2019.00027
M3 - Conference contribution
AN - SCOPUS:85074656794
T3 - Proceedings of the IEEE International Conference on Engineering of Complex Computer Systems, ICECCS
SP - 180
EP - 189
BT - Proceedings - 2019 24th International Conference on Engineering of Complex Computer Systems, ICECCS 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 24th International Conference on Engineering of Complex Computer Systems, ICECCS 2019
Y2 - 10 November 2019 through 13 November 2019
ER -