4.7 Article

Generating probabilistic safety guarantees for neural network controllers

Related references

Note: Only part of the references are listed.
Article Automation & Control Systems

Formal verification of neural agents in non-deterministic environments

Michael E. Akintunde et al.

Summary: This paper introduces a model for agent-environment systems and investigates the verification problem against CTL properties. It shows that verifying reachability properties in these systems is undecidable. The paper also introduces a bounded fragment of CTL and presents sequential and parallel algorithms for verifying agent-environment systems using MILP. Experimental results are reported against different use-cases.

AUTONOMOUS AGENTS AND MULTI-AGENT SYSTEMS (2022)

Proceedings Paper Computer Science, Hardware & Architecture

Safety Verification of Neural Network Controlled Systems

Arthur Claviere et al.

Summary: This paper introduces a system-level approach for verifying the safety of systems that combine continuous-time physical systems with discrete-time neural network controllers. The approach includes a generic modeling method and reachability analysis that accurately approximates reachable states of the entire system, demonstrated through a real-world use case.

51ST ANNUAL IEEE/IFIP INTERNATIONAL CONFERENCE ON DEPENDABLE SYSTEMS AND NETWORKS (DSN-W 2021) (2021)

Article Computer Science, Hardware & Architecture

ReachNN: Reachability Analysis of Neural-Network Controlled Systems

Chao Huang et al.

ACM TRANSACTIONS ON EMBEDDED COMPUTING SYSTEMS (2019)

Article Engineering, Aerospace

Deep Neural Network Compression for Aircraft Collision Avoidance Systems

Kyle D. Julian et al.

JOURNAL OF GUIDANCE CONTROL AND DYNAMICS (2019)

Proceedings Paper Computer Science, Software Engineering

The Marabou Framework for Verification and Analysis of Deep Neural Networks

Guy Katz et al.

COMPUTER AIDED VERIFICATION, CAV 2019, PT I (2019)

Proceedings Paper Computer Science, Information Systems

Reachability Analysis for Neural Feedback Systems using Regressive Polynomial Rule Inference

Souradeep Dutta et al.

PROCEEDINGS OF THE 2019 22ND ACM INTERNATIONAL CONFERENCE ON HYBRID SYSTEMS: COMPUTATION AND CONTROL (HSCC '19) (2019)

Proceedings Paper Computer Science, Information Systems

Verisig: verifying safety properties of hybrid systems with neural network controllers

Radoslav Ivanov et al.

PROCEEDINGS OF THE 2019 22ND ACM INTERNATIONAL CONFERENCE ON HYBRID SYSTEMS: COMPUTATION AND CONTROL (HSCC '19) (2019)

Proceedings Paper Computer Science, Theory & Methods

Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networks

Guy Katz et al.

COMPUTER AIDED VERIFICATION, CAV 2017, PT I (2017)

Article Multidisciplinary Sciences

Human-level control through deep reinforcement learning

Volodymyr Mnih et al.

NATURE (2015)

Article Computer Science, Artificial Intelligence

Variable resolution discretization in optimal control

R Munos et al.

MACHINE LEARNING (2002)