Advertisement

Repairing Learned Controllers with Convex Optimization: A Case Study

  • Dario GuidottiEmail author
  • Francesco Leofante
  • Claudio Castellini
  • Armando Tacchella
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11494)

Abstract

Despite the increasing popularity of Machine Learning methods, their usage in safety-critical applications is sometimes limited by the impossibility of providing formal guarantees on their behaviour. In this work we focus on one such application, where Kernel Ridge Regression with Random Fourier Features is used to learn controllers for a prosthetic hand. Due to the non-linearity of the activation function used, these controllers sometimes fail in correctly identifying users’ intention. Under specific circumstances muscular activation levels may be misinterpreted by the method, resulting in the prosthetic hand not behaving as intended. To alleviate this problem, we propose a novel method to verify the presence of this kind of intent detection mismatch and to repair controllers leveraging off-the-shelf LP technology without using additional data. We demonstrate the feasibility of our approach using datasets gathered from human participants.

References

  1. 1.
    Taigman, Y., Yang, M., Ranzato, M., Wolf, L.: DeepFace: closing the gap to human-level performance in face verification. In: CVPR, pp. 1701–1708 (2014)Google Scholar
  2. 2.
    Yu, D., Hinton, G.E., Morgan, N., Chien, J.-T., Sagayama, S.: Introduction to the special section on deep learning for speech and language processing. IEEE Trans. Audio Speech Lang. Process. 20(1), 4–6 (2012)CrossRefGoogle Scholar
  3. 3.
    LeCun, Y., Bengio, Y., Hinton, G.E.: Deep learning. Nature 521(7553), 436–444 (2015)CrossRefGoogle Scholar
  4. 4.
    Rahimi, A., Recht, B.: Random features for large-scale kernel machines. In: NIPS, pp. 1177–1184 (2008)Google Scholar
  5. 5.
    Gijsberts, A., Metta, G.: Incremental learning of robot dynamics using random features. In: ICRA, pp. 951–956 (2011)Google Scholar
  6. 6.
    Fougner, A., Stavdahl, Ø., Kyberd, P.J., Losier, Y.G., Parker, P.A.: Control of upper limb prostheses: terminology and proportional myoelectric control - a review. IEEE Trans. Neural Syst. Rehabil. Eng. 20(5), 663–677 (2012)CrossRefGoogle Scholar
  7. 7.
    Merletti, R., Botter, A., Cescon, C., Minetto, M.A., Vieira, T.M.M.: Advances in surface EMG: recent progress in clinical research applications. Crit. Rev. Biomed. Eng. 38(4), 347–379 (2011)CrossRefGoogle Scholar
  8. 8.
    Gijsberts, A., et al.: Stable myoelectric control of a hand prosthesis using non-linear incremental learning. Front. Neurorobot. 8 (2014)Google Scholar
  9. 9.
    Strazzulla, I., Nowak, M., Controzzi, M., Cipriani, C., Castellini, C.: Online bimanual manipulation using surface electromyography and incremental learning. IEEE Trans. Neural Syst. Rehabil. Eng. 25(3), 227–234 (2017)CrossRefGoogle Scholar
  10. 10.
    Gestel, T.V., et al.: Benchmarking least squares support vector machine classifiers. Mach. Learn. 54(1), 5–32 (2004)CrossRefGoogle Scholar
  11. 11.
    Patel, G., Nowak, M., Castellini, C.: Exploiting knowledge composition to improve real-life hand prosthetic control. IEEE Trans. Neural Syst. Rehabil. Eng. 25(7), 967–975 (2017)CrossRefGoogle Scholar
  12. 12.
    Schrijver, A.: Theory of Linear and Integer Programming. Wiley-Interscience Series in Discrete Mathematics and Optimization. Wiley, Hoboken (1999)zbMATHGoogle Scholar
  13. 13.
    Leofante, F., Narodytska, N., Pulina, L., Tacchella, A.: Automated verification of neural networks: advances, challenges and perspectives. arXiv preprint arXiv:1805.09938 (2018)
  14. 14.
    Narodytska, N., Kasiviswanathan, S.P., Ryzhyk, L., Sagiv, M., Walsh, T.: Verifying properties of binarized deep neural networks. In: AAAI, pp. 6615–6624 (2018)Google Scholar
  15. 15.
    Pulina, L., Tacchella, A.: An abstraction-refinement approach to verification of artificial neural networks. In: Touili, T., Cook, B., Jackson, P. (eds.) CAV 2010. LNCS, vol. 6174, pp. 243–257. Springer, Heidelberg (2010).  https://doi.org/10.1007/978-3-642-14295-6_24CrossRefGoogle Scholar
  16. 16.
    Katz, G., Barrett, C., Dill, D.L., Julian, K., Kochenderfer, M.J.: Reluplex: an efficient SMT solver for verifying deep neural networks. In: Majumdar, R., Kunčak, V. (eds.) CAV 2017. LNCS, vol. 10426, pp. 97–117. Springer, Cham (2017).  https://doi.org/10.1007/978-3-319-63387-9_5CrossRefGoogle Scholar
  17. 17.
    Leofante, F., Tacchella, A.: Learning in physical domains: mating safety requirements and costly sampling. In: Adorni, G., Cagnoni, S., Gori, M., Maratea, M. (eds.) AI*IA 2016. LNCS (LNAI), vol. 10037, pp. 539–552. Springer, Cham (2016).  https://doi.org/10.1007/978-3-319-49130-1_39CrossRefGoogle Scholar
  18. 18.
    Cheng, C.-H., Nührenberg, G., Ruess, H.: Maximum resilience of artificial neural networks. In: D’Souza, D., Narayan Kumar, K. (eds.) ATVA 2017. LNCS, vol. 10482, pp. 251–268. Springer, Cham (2017).  https://doi.org/10.1007/978-3-319-68167-2_18CrossRefGoogle Scholar
  19. 19.
    Fischetti, M., Jo, J.: Deep neural networks and mixed integer linear optimization. Constraints 23(3), 296–309 (2018)MathSciNetCrossRefGoogle Scholar
  20. 20.
    Ehlers, R.: Formal verification of piece-wise linear feed-forward neural networks. In: D’Souza, D., Narayan Kumar, K. (eds.) ATVA 2017. LNCS, vol. 10482, pp. 269–286. Springer, Cham (2017).  https://doi.org/10.1007/978-3-319-68167-2_19CrossRefGoogle Scholar
  21. 21.
    Gehr, T., Mirman, M., Drachsler-Cohen, D., Tsankov, P., Chaudhuri, S., Vechev, M.: AI2: safety and robustness certification of neural networks with abstract interpretation. In: 2018 IEEE Symposium on Security and Privacy (SP) (2018)Google Scholar
  22. 22.
    Valero-Cuevas, F.J.: Predictive modulation of muscle coordination pattern magnitude scales fingertip force magnitude over the voluntary range. J. Neurophysiol. 83(3), 1469–1479 (2000)CrossRefGoogle Scholar
  23. 23.
    Poston, B., Danna-Dos Santos, A., Jesunathadas, M., Hamm, T.M., Santello, M.: Force-independent distribution of correlated neural inputs to hand muscles during three-digit grasping. J. Neurophysiol. 104(2), 1141–1154 (2010)CrossRefGoogle Scholar
  24. 24.
    de Rugy, A., Loeb, G.E., Carroll, T.J.: Muscle coordination is habitual rather than optimal. J. Neurosci. 32(21), 7384–7391 (2012)CrossRefGoogle Scholar
  25. 25.
    He, J., Zhang, D., Sheng, X., Li, S., Zhu, X.: Invariant surface emg feature against varying contraction level for myoelectric control based on muscle coordination. IEEE J. Biomed. Heal. Inform. 19(3), 874–882 (2015)Google Scholar
  26. 26.
    Al-Timemy, A.H., Khushaba, R.N., Bugmann, G., Escudero, J.: Improving the performance against force variation of emg controlled multifunctional upper-limb prostheses for transradial amputees. IEEE Trans. Neural Syst. Rehabil. Eng. 24(6), 650–661 (2016)CrossRefGoogle Scholar
  27. 27.
    Vandenberghe, L.: The CVXOPT linear and quadratic cone program solvers (2010)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Dario Guidotti
    • 1
    Email author
  • Francesco Leofante
    • 1
  • Claudio Castellini
    • 2
  • Armando Tacchella
    • 1
  1. 1.DIBRISUniversità degli Studi di GenovaGenoaItaly
  2. 2.Institute of Robotics and Mechatronics, German Aerospace CenterWeßlingGermany

Personalised recommendations