Promises of Deep Kernel Learning for Control Synthesis

被引:2
|
作者
Reed, Robert [1 ]
Laurenti, Luca [2 ]
Lahijanian, Morteza [1 ]
机构
[1] Univ Colorado Boulder, Dept Aerosp Engn Sci, Boulder, CO 80304 USA
[2] Delft Univ Technol, Delft Ctr Syst & Control, NL-2628 CN Delft, Netherlands
来源
IEEE CONTROL SYSTEMS LETTERS | 2023年 / 7卷
关键词
Machine learning; robust control; stochastic systems; VERIFICATION;
D O I
10.1109/LCSYS.2023.3340995
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Deep Kernel Learning (DKL) combines the representational power of neural networks with the uncertainty quantification of Gaussian Processes. Hence, it is potentially a promising tool to learn and control complex dynamical systems. In this letter, we develop a scalable abstraction-based framework that enables the use of DKL for control synthesis of stochastic dynamical systems against complex specifications. Specifically, we consider temporal logic specifications and create an end-to-end framework that uses DKL to learn an unknown system from data and formally abstracts the DKL model into an interval Markov decision process to perform control synthesis with correctness guarantees. Furthermore, we identify a deep architecture that enables accurate learning and efficient abstraction computation. The effectiveness of our approach is illustrated on various benchmarks, including a 5-D nonlinear stochastic system, showing how control synthesis with DKL can substantially outperform state-of-the-art competitive methods.
引用
收藏
页码:3986 / 3991
页数:6
相关论文
共 50 条
  • [1] Bridging deep and multiple kernel learning: A review
    Wang, Tinghua
    Zhang, Lin
    Hu, Wenyu
    INFORMATION FUSION, 2021, 67 : 3 - 13
  • [2] Computational biology perspective: kernel methods and deep learning
    Lodhi, Huma
    WILEY INTERDISCIPLINARY REVIEWS-COMPUTATIONAL STATISTICS, 2012, 4 (05): : 455 - 465
  • [3] The Quantum Path Kernel: A Generalized Neural Tangent Kernel for Deep Quantum Machine Learning
    Incudini M.
    Grossi M.
    Mandarino A.
    Vallecorsa S.
    Pierro A.D.
    Windridge D.
    IEEE Transactions on Quantum Engineering, 2023, 4
  • [4] Energy Commodity Price Forecasting with Deep Multiple Kernel Learning
    Huang, Shian-Chang
    Wu, Cheng-Feng
    ENERGIES, 2018, 11 (11)
  • [5] Kernel Mapping Techniques for Deep Learning Neural Network Accelerators
    Ozdemir, Sarp
    Khasawneh, Mohammad
    Rao, Smriti
    Madden, Patrick H.
    ISPD'22: PROCEEDINGS OF THE 2022 INTERNATIONAL SYMPOSIUM ON PHYSICAL DESIGN, 2022, : 21 - 28
  • [6] Neural Kernel Network Deep Kernel Learning for Predicting Particulate Matter from Heterogeneous Sensors with Uncertainty
    Li, Chaofan
    Riedel, Till
    Beigl, Michael
    INFORMATION INTEGRATION AND WEB INTELLIGENCE, IIWAS 2022, 2022, 13635 : 252 - 266
  • [7] A predictive intelligence system of credit scoring based on deep multiple kernel learning
    Wu, Cheng-Feng
    Huang, Shian-Chang
    Chiou, Chei-Chang
    Wang, Yu-Min
    APPLIED SOFT COMPUTING, 2021, 111 (111)
  • [8] Leveraging deep learning to control neural oscillators
    Timothy D. Matchen
    Jeff Moehlis
    Biological Cybernetics, 2021, 115 : 219 - 235
  • [9] Optimal deep learning control for modernized microgrids
    Yan, Shu-Rong
    Guo, Wei
    Mohammadzadeh, Ardashir
    Rathinasamy, Sakthivel
    APPLIED INTELLIGENCE, 2023, 53 (12) : 15638 - 15655
  • [10] Deep reinforcement learning for inventory control: A roadmap
    Boute, Robert N.
    Gijsbrechts, Joren
    van Jaarsveld, Willem
    Vanvuchelen, Nathalie
    EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2022, 298 (02) : 401 - 412