NNTBFV: Simplifying and Verifying Neural Networks Using Testing-Based Formal Verification

被引:1
作者
Liu, Haiyi [1 ]
Liu, Shaoying [1 ]
Xu, Guangquan [2 ]
Liu, Ai [1 ]
Fang, Dingbang [1 ]
机构
[1] Hiroshima Univ, Grad Sch Adv Sci & Engn, Hiroshima 7398527, Japan
[2] Tianjin Univ, Coll Intelligence & Comp, Sch Cybersecur, Tianjin 300072, Peoples R China
关键词
Neural network simplification; formal verification; neural network reliability;
D O I
10.1142/S0218194023500523
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Neural networks are extensively employed in safety-critical systems. However, these critical systems incorporating neural networks continue to pose risks due to the presence of adversarial examples. Although the security of neural networks can be enhanced by verification, verifying neural networks is an NP-hard problem, making the application of verification algorithms to large-scale neural networks a challenging task. For this reason, we propose NNTBFV, a framework that utilizes the principles of Testing-Based Formal Verification (TBFV) to simplify neural networks and verify the simplified networks. Unlike conventional neural network pruning techniques, this approach is based on specifications, with the goal of deriving approximate execution paths under given preconditions. To mitigate the potential issue of unverifiable conditions due to overly broad preconditions, we also propose a precondition partition method. Empirical evidence shows that as the range of preconditions narrows, the size of the execution paths also reduces accordingly. The execution path generated by NNTBFV is still a neural network, so it can be verified by verification tools. In response to the results from the verification tool, we provide a theoretical method for analysis.We evaluate the effectiveness of NNTBFV on the ACAS Xu model project, choosing Verification-based and Random-based neural network simplification algorithms as the baselines for NNTBFV. Experiment results show that NNTBFV can effectively approximate the baseline in terms of simplification capability, and it surpasses the efficiency of the random-based method.
引用
收藏
页码:273 / 300
页数:28
相关论文
共 32 条
  • [21] Formal verification of Matrix based MATLAB models using interactive theorem proving
    Gauhar, Ayesha
    Rashid, Adnan
    Hasan, Osman
    Bispo, Joao
    Cardoso, Joao M. P.
    PEERJ COMPUTER SCIENCE, 2021, 7 : 1 - 21
  • [22] An Efficient Approach Towards Formal Verification of Mixed Signals Using Feed-Forward Neural Network
    Vidhya, D. S.
    Ramachandra, Manjunath
    CYBERNETICS AND AUTOMATION CONTROL THEORY METHODS IN INTELLIGENT ALGORITHMS, 2019, 986 : 31 - 39
  • [23] Formal verification of IoT applications using rewriting logic: An MDE-based approach
    Fortas, Abdelouahab
    Kerkouche, Elhillali
    Chaoui, Allaoua
    SCIENCE OF COMPUTER PROGRAMMING, 2022, 222
  • [24] Adopting Formal Verification and Model-Based Testing Techniques for Validating a Blockchain-based Healthcare Records Sharing System
    Jabbar, Rateb
    Krichen, Moez
    Fetais, Noora
    Barkaoui, Kamel
    PROCEEDINGS OF THE 22ND INTERNATIONAL CONFERENCE ON ENTERPRISE INFORMATION SYSTEMS (ICEIS), VOL 1, 2020, : 261 - 268
  • [25] NNLander-VeriF: A Neural Network Formal Verification Framework for Vision-Based Autonomous Aircraft Landing
    Santa Cruz, Ulices
    Shoukry, Yasser
    NASA FORMAL METHODS (NFM 2022), 2022, 13260 : 213 - 230
  • [26] Formal Specification and Verification of Smart Contract-Based Loan Management System Using TLA
    Yoon, Seongho
    Choi, Jin-Young
    IEEE ACCESS, 2025, 13 : 62060 - 62070
  • [27] An improved genetic algorithm for task scheduling in the cloud environments using the priority queues: Formal verification, simulation, and statistical testing
    Keshanchi, Bahman
    Souri, Alireza
    Navimipour, Nima Jafari
    JOURNAL OF SYSTEMS AND SOFTWARE, 2017, 124 : 1 - 21
  • [28] Groebner basis based formal verification of large arithmetic circuits using Gaussian elimination and cone-based polynomial extraction
    Farahmandi, Farimah
    Alizadeh, Bijan
    MICROPROCESSORS AND MICROSYSTEMS, 2015, 39 (02) : 83 - 96
  • [29] Intrusion Detection in PLC-Based Industrial Control Systems Using Formal Verification Approach in Conjunction with Graphs
    Muluken Hailesellasie
    Syed Rafay Hasan
    Journal of Hardware and Systems Security, 2018, 2 (1) : 1 - 14
  • [30] Using Formal Specifications to Support Model Based Testing ASDSpec: A Tool Combining the Best of Two Techniques
    van der Meer, A. P.
    Kherrazi, R.
    Hamilton, M.
    ELECTRONIC PROCEEDINGS IN THEORETICAL COMPUTER SCIENCE, 2013, (141): : 1 - 13