Back to MJ's Publications

Publications of year 2022
Theses
  1. D. Ding. Provable reinforcement learning for constrained and multi-agent control systems. PhD thesis, University of Southern California, 2022. Keyword(s): Constrained Markov decision processes, Constrained nonconvex optimization, Function approximation, Game-agnostic convergence, Multi-agent reinforcement learning, Multi-agent systems, Natural policy gradient, Policy gradient methods, Proximal policy optimization, Primal-dual algorithms, Reinforcement learning, Safe exploration, Safe reinforcement learning, Sample complexity, Stochastic optimization.
    @PHDTHESIS{dongsheng-phd22,
    AUTHOR = {D. Ding},
    SCHOOL = {University of Southern California},
    TITLE = {Provable reinforcement learning for constrained and multi-agent control systems},
    YEAR = {2022},
    KEYWORDS = {Constrained Markov decision processes, Constrained nonconvex optimization, Function approximation, Game-agnostic convergence, Multi-agent reinforcement learning, Multi-agent systems, Natural policy gradient, Policy gradient methods, Proximal policy optimization, Primal-dual algorithms, Reinforcement learning, Safe exploration, Safe reinforcement learning, Sample complexity, Stochastic optimization},
    PDF = {https://viterbi-web.usc.edu/~mihailo/theses/DongshengDing-PhD.pdf} 
    }
    


  2. H. Mohammadi. Robustness of gradient methods for data-driven decision making. PhD thesis, University of Southern California, 2022. Keyword(s): Accelerated first-order algorithms, Control for optimization, Convergence rate, Convex optimization, Data-driven control, Gradient descent, Gradient-flow dynamics, Heavy-ball method, Integral quadratic constraints, Linear quadratic regulator, Model-free control, Nesterov's accelerated method, Nonconvex optimization, Nonnormal dynamics, Noise amplification, Optimization, Optimal control, Polyak-Lojasiewicz inequality, Random search method, Reinforcement learning, Sample complexity, Second-order moments, Transient growth.
    @PHDTHESIS{hesam-phd22,
    AUTHOR = {H. Mohammadi},
    SCHOOL = {University of Southern California},
    TITLE = {Robustness of gradient methods for data-driven decision making},
    YEAR = {2022},
    KEYWORDS = {Accelerated first-order algorithms, Control for optimization, Convergence rate, Convex optimization, Data-driven control, Gradient descent, Gradient-flow dynamics, Heavy-ball method, Integral quadratic constraints, Linear quadratic regulator, Model-free control, Nesterov's accelerated method, Nonconvex optimization, Nonnormal dynamics, Noise amplification, Optimization, Optimal control, Polyak-Lojasiewicz inequality, Random search method, Reinforcement learning, Sample complexity, Second-order moments, Transient growth},
    PDF = {https://viterbi-web.usc.edu/~mihailo/theses/HesammedinMohammadi-PhD.pdf} 
    }
    


Journal articles
  1. H. A. Castillo, M. R. Jovanovic, S. Kumar, A. Morozov, V. Shankar, G. Subramanian, and H. J. Wilson. Understanding viscoelastic flow instabilities: Oldroyd-B and beyond. J. Non-Newtonian Fluid Mech., 302:104742 (39 pages), April 2022. Note: Part of the special issue commemorating the birth centenary of James Oldroyd. Keyword(s): Flow modeling and control, Input-output analysis, Elastic turbulence, Energy amplification, Transition to turbulence, Uncertainty quantification in PDEs, Spatio-temporal frequency responses, Viscoelastic fluids, Viscoelastic instabilities.
    @article{oldroyd100,
    author = {H. A. Castillo and M. R. Jovanovi\'c and S. Kumar and A. Morozov and V. Shankar and G. Subramanian and H. J. Wilson},
    title = {Understanding viscoelastic flow instabilities: {O}ldroyd-{B} and beyond},
    journal = {J. Non-Newtonian Fluid Mech.},
    volume = {302},
    pages = {104742 (39 pages)},
    month = {April},
    year = {2022},
    note = {{P}art of the special issue commemorating the birth centenary of {J}ames {O}ldroyd},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/oldroyd100.pdf},
    keywords = {Flow modeling and control, Input-output analysis, Elastic turbulence, Energy amplification, Transition to turbulence, Uncertainty quantification in PDEs, Spatio-temporal frequency responses, Viscoelastic fluids, Viscoelastic instabilities} 
    }
    


  2. N. K. Dhingra, S. Z. Khong, and M. R. Jovanovic. A second order primal-dual method for nonsmooth convex composite optimization. IEEE Trans. Automat. Control, 67(8):4061-4076, August 2022. Keyword(s): Augmented Lagrangian, Exponential convergence, Global exponential stability, Method of multipliers, Non-smooth optimization, Proximal algorithms, Proximal augmented Lagrangian, Regularization for design, Second order primal-dual method, Sparsity-promoting optimal control, Structured optimal control, Structure identification.
    @ARTICLE{dhikhojovTAC22,
    AUTHOR = {N. K. Dhingra and S. Z. Khong and M. R. Jovanovi\'c},
    TITLE = {A second order primal-dual method for nonsmooth convex composite optimization},
    JOURNAL = {IEEE Trans. Automat. Control},
    VOLUME = {67},
    NUMBER = {8},
    PAGES = {4061-4076},
    MONTH = {August},
    YEAR = {2022},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/dhikhojovTAC22.pdf},
    KEYWORDS = {Augmented Lagrangian, Exponential convergence, Global exponential stability, Method of multipliers, Non-smooth optimization, Proximal algorithms, Proximal augmented Lagrangian, Regularization for design, Second order primal-dual method, Sparsity-promoting optimal control, Structured optimal control, Structure identification} 
    }
    


  3. D. Ding, K. Zhang, J. Duan, T. Basar, and M. R. Jovanovic. Convergence and sample complexity of natural policy gradient primal-dual methods for constrained MDPs. J. Mach. Learn. Res., 2022. Note: Submitted; also arXiv:2206.02346. Keyword(s): Constrained Markov decision processes, Constrained nonconvex optimization, Function approximation, Natural policy gradient, Policy gradient methods, Primal-dual algorithms, Sample complexity.
    @ARTICLE{dinzhaduabasjovJMLR22,
    AUTHOR = {D. Ding and K. Zhang and J. Duan and T. Basar and M. R. Jovanovi\'c},
    TITLE = {Convergence and sample complexity of natural policy gradient primal-dual methods for constrained {MDP}s },
    JOURNAL = {J. Mach. Learn. Res.},
    YEAR = {2022},
    NOTE = {submitted; also arXiv:2206.02346},
    PDF = {https://arxiv.org/abs/2206.02346},
    KEYWORDS = {Constrained Markov decision processes, Constrained nonconvex optimization, Function approximation, Natural policy gradient, Policy gradient methods, Primal-dual algorithms, Sample complexity} 
    }
    


  4. A. Dwivedi, G. S. Sidharth, and M. R. Jovanovic. Oblique transition in hypersonic double-wedge flow. J. Fluid Mech., 948:A37, October 2022. Keyword(s): Compressible flows, Direct numerical simulations, Double-wedge flow, Flow modeling and control, Hypersonic flows, Input-output analysis, Oblique waves, Shock boundary layer interaction, Reattachment vortices, Transition to turbulence, Weakly nonlinear analysis.
    @ARTICLE{dwisidjovJFM22,
    AUTHOR = {A. Dwivedi and G. S. Sidharth and M. R. Jovanovi\'c},
    JOURNAL = {J. Fluid Mech.},
    TITLE = {Oblique transition in hypersonic double-wedge flow},
    VOLUME = {948},
    PAGES = {A37},
    MONTH = {October},
    YEAR = {2022},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/dwisidjovJFM22.pdf},
    KEYWORDS = {Compressible flows, Direct numerical simulations, Double-wedge flow, Flow modeling and control, Hypersonic flows, Input-output analysis, Oblique waves, Shock boundary layer interaction, Reattachment vortices, Transition to turbulence, Weakly nonlinear analysis} 
    }
    


  5. H. Mohammadi, M. Razaviyayn, and M. R. Jovanovic. Tradeoffs between convergence rate and noise amplification for momentum-based accelerated optimization algorithms. IEEE Trans. Automat. Control, 2022. Note: Submitted; also arXiv:2209.11920. Keyword(s): Accelerated first-order algorithms, Control for optimization, Convergence rate, Convex optimization, Gradient descent, Heavy-ball method, Nesterov's accelerated method, Nonnormal dynamics, Noise amplification, Second-order moments.
    @ARTICLE{mohrazjovTAC22,
    AUTHOR = {H. Mohammadi and M. Razaviyayn and M. R. Jovanovi\'c},
    TITLE = {Tradeoffs between convergence rate and noise amplification for momentum-based accelerated optimization algorithms},
    JOURNAL = {IEEE Trans. Automat. Control},
    YEAR = {2022},
    NOTE = {submitted; also arXiv:2209.11920},
    PDF = {https://arxiv.org/abs/2209.11920},
    KEYWORDS = {Accelerated first-order algorithms, Control for optimization, Convergence rate, Convex optimization, Gradient descent, Heavy-ball method, Nesterov's accelerated method, Nonnormal dynamics, Noise amplification, Second-order moments} 
    }
    


  6. H. Mohammadi, A. Zare, M. Soltanolkotabi, and M. R. Jovanovic. Convergence and sample complexity of gradient methods for the model-free linear-quadratic regulator problem. IEEE Trans. Automat. Control, 67(5):2435-2450, May 2022. Keyword(s): Data-driven control, Gradient descent, Gradient-flow dynamics, Linear quadratic regulator, Model-free control, Nonconvex optimization, Optimization, Optimal control, Polyak-Lojasiewicz inequality, Random search method, Reinforcement learning, Sample complexity.
    @ARTICLE{mohzarsoljovTAC22,
    AUTHOR = {H. Mohammadi and A. Zare and M. Soltanolkotabi and M. R. Jovanovi\'c},
    TITLE = {Convergence and sample complexity of gradient methods for the model-free linear-quadratic regulator problem},
    JOURNAL = {IEEE Trans. Automat. Control},
    VOLUME = {67},
    NUMBER = {5},
    PAGES = {2435-2450},
    MONTH = {May},
    YEAR = {2022},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/mohzarsoljovTAC22.pdf},
    KEYWORDS = {Data-driven control, Gradient descent, Gradient-flow dynamics, Linear quadratic regulator, Model-free control, Nonconvex optimization, Optimization, Optimal control, Polyak-Lojasiewicz inequality, Random search method, Reinforcement learning, Sample complexity} 
    }
    


Conference articles
  1. L. Ballotta, M. R. Jovanovic, and L. Schenato. Can decentralized control outperform centralized? The role of communication latency. In Proceedings of the 2022 IFAC Conference on Networked Systems, Zurich, Switzerland, 2022. Keyword(s): Controller architecture, Fundamental limitations, Networks, Networks of dynamical systems, Noise amplification, Performance bounds, Topology design.
    @INPROCEEDINGS{baljovschIFAC22,
    AUTHOR = {L. Ballotta and M. R. Jovanovi\'c and L. Schenato},
    TITLE = {Can decentralized control outperform centralized? {T}he role of communication latency},
    BOOKTITLE = {Proceedings of the 2022 IFAC Conference on Networked Systems},
    YEAR = {2022},
    ADDRESS = {Zurich, Switzerland},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/baljovschIFAC22.pdf},
    KEYWORDS = {Controller architecture, Fundamental limitations, Networks, Networks of dynamical systems, Noise amplification, Performance bounds, Topology design} 
    }
    


  2. D. Ding and M. R. Jovanovic. Policy gradient primal-dual mirror descent for constrained MDPs with large state spaces. In Proceedings of the 61st IEEE Conference on Decision and Control, Cancun, Mexico, pages 4892-4897, 2022. Keyword(s): Constrained Markov decision processes, Policy gradient methods, Primal-dual algorithms, Mirror descent, Function approximation.
    @INPROCEEDINGS{dinjovCDC22,
    AUTHOR = {D. Ding and M. R. Jovanovi\'c},
    BOOKTITLE = {Proceedings of the 61st IEEE Conference on Decision and Control},
    TITLE = {Policy gradient primal-dual mirror descent for constrained {MDP}s with large state spaces},
    YEAR = {2022},
    ADDRESS = {Cancun, Mexico},
    PAGES = {4892-4897},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/dinjovCDC22.pdf},
    KEYWORDS = {Constrained Markov decision processes, Policy gradient methods, Primal-dual algorithms, Mirror descent, Function approximation} 
    }
    


  3. D. Ding, C.-Y. Wei, K. Zhang, and M. R. Jovanovic. Independent policy gradient for large-scale Markov potential games: sharper rates, function approximation, and game-agnostic convergence. In Proceedings of the 39th International Conference on Machine Learning, volume 162 of Proceedings of Machine Learning Research, Baltimore, MD, pages 5166-5220, 2022. Keyword(s): Multi-agent reinforcement learning, Independent reinforcement learning, Policy gradient methods, Markov potential games, Function approximation, Game-agnostic convergence.
    @INPROCEEDINGS{dinweizhajovICML22,
    AUTHOR = {D. Ding and C.-Y. Wei and K. Zhang and M. R. Jovanovi\'c},
    TITLE = {Independent policy gradient for large-scale {M}arkov potential games: sharper rates, function approximation, and game-agnostic convergence},
    BOOKTITLE = {Proceedings of the 39th International Conference on Machine Learning},
    SERIES = {Proceedings of Machine Learning Research},
    VOLUME = {162},
    PAGES = {5166-5220},
    YEAR = {2022},
    ADDRESS = {Baltimore, MD},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/dinweizhajovICML22.pdf},
    KEYWORDS = {Multi-agent reinforcement learning, Independent reinforcement learning, Policy gradient methods, Markov potential games, Function approximation, Game-agnostic convergence} 
    }
    


  4. D. Ding, K. Zhang, T. Basar, and M. R. Jovanovic. Convergence and optimality of policy gradient primal-dual method for constrained Markov decision processes. In Proceedings of the 2022 American Control Conference, Atlanta, GA, pages 2851-2856, 2022. Keyword(s): Constrained Markov decision processes, Policy gradient methods, Primal-dual algorithms.
    @INPROCEEDINGS{dinzhabasjovACC22,
    AUTHOR = {D. Ding and K. Zhang and T. Basar and M. R. Jovanovi\'c},
    TITLE = {Convergence and optimality of policy gradient primal-dual method for constrained {M}arkov decision processes},
    BOOKTITLE = {Proceedings of the 2022 American Control Conference},
    PAGES = {2851-2856},
    YEAR = {2022},
    ADDRESS = {Atlanta, GA},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/dinzhabasjovACC22.pdf},
    KEYWORDS = {Constrained Markov decision processes, Policy gradient methods, Primal-dual algorithms} 
    }
    


  5. A. Dwivedi and M. R. Jovanovic. A weakly nonlinear analysis of transition in a hypersonic flow. In Proceedings of the 2022 American Control Conference, Atlanta, GA, pages 4171-4176, 2022. Keyword(s): Compressible flows, Direct numerical simulations, Double-wedge flow, Flow modeling and control, Hypersonic flows, Input-output analysis, Shock boundary layer interaction, Reattachment vortices, Transition to turbulence, Weakly nonlinear analysis.
    @INPROCEEDINGS{dwijovACC22,
    AUTHOR = {A. Dwivedi and M. R. Jovanovi\'c},
    TITLE = {A weakly nonlinear analysis of transition in a hypersonic flow},
    BOOKTITLE = {Proceedings of the 2022 American Control Conference},
    PAGES = {4171-4176},
    YEAR = {2022},
    ADDRESS = {Atlanta, GA},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/dwijovACC22.pdf},
    KEYWORDS = {Compressible flows, Direct numerical simulations, Double-wedge flow, Flow modeling and control, Hypersonic flows, Input-output analysis, Shock boundary layer interaction, Reattachment vortices, Transition to turbulence, Weakly nonlinear analysis} 
    }
    


  6. H. Mohammadi and M. R. Jovanovic. On the noise amplification of primal-dual gradient flow dynamics based on proximal augmented Lagrangian. In Proceedings of the 2022 American Control Conference, Atlanta, GA, pages 926-931, 2022. Keyword(s): Control for optimization, Convex Optimization, Integral quadratic constraints, Linear matrix inequalities, Noise amplification, Non-smooth optimization, Proximal algorithms, Primal-dual gradient flow dynamics, Primal-dual methods, Proximal augmented Lagrangian, Second-order moments, Semidefinite programming.
    @INPROCEEDINGS{mohjovACC22,
    AUTHOR = {H. Mohammadi and M. R. Jovanovi\'c},
    TITLE = {On the noise amplification of primal-dual gradient flow dynamics based on proximal augmented {L}agrangian},
    BOOKTITLE = {Proceedings of the 2022 American Control Conference},
    PAGES = {926-931},
    YEAR = {2022},
    ADDRESS = {Atlanta, GA},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/mohjovACC22.pdf},
    KEYWORDS = {Control for optimization, Convex Optimization, Integral quadratic constraints, Linear matrix inequalities, Noise amplification, Non-smooth optimization, Proximal algorithms, Primal-dual gradient flow dynamics, Primal-dual methods, Proximal augmented Lagrangian, Second-order moments, Semidefinite programming} 
    }
    


  7. I. K. Ozaslan, S. Hassan-Moghaddam, and M. R. Jovanovic. On the asymptotic stability of proximal algorithms for convex optimization problems with multiple non-smooth regularizers. In Proceedings of the 2022 American Control Conference, Atlanta, GA, pages 132-137, 2022. Keyword(s): Control for optimization, Convex Optimization, Douglas-Rachford splitting, Global asymptotic stability, Lyapunov-based analysis, Non-smooth optimization, Proximal algorithms, Primal-dual gradient flow dynamics, Primal-dual methods, Proximal augmented Lagrangian.
    @INPROCEEDINGS{ozamogjovACC22,
    AUTHOR = {I. K. Ozaslan and S. Hassan-Moghaddam and M. R. Jovanovi\'c},
    TITLE = {On the asymptotic stability of proximal algorithms for convex optimization problems with multiple non-smooth regularizers},
    BOOKTITLE = {Proceedings of the 2022 American Control Conference},
    PAGES = {132-137},
    YEAR = {2022},
    ADDRESS = {Atlanta, GA},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/ozamogjovACC22.pdf},
    KEYWORDS = {Control for optimization, Convex Optimization, Douglas-Rachford splitting, Global asymptotic stability, Lyapunov-based analysis, Non-smooth optimization, Proximal algorithms, Primal-dual gradient flow dynamics, Primal-dual methods, Proximal augmented Lagrangian} 
    }
    


  8. I. K. Ozaslan and M. R. Jovanovic. Exponential convergence of primal-dual dynamics for multi-block problems under local error bound condition. In Proceedings of the 61th IEEE Conference on Decision and Control, Cancun, Mexico, pages 7579-7584, 2022. Keyword(s): Gradient flow dynamics, Lyapunov functions, Proximal algorithms, Primal-dual gradient flow dynamics, Primal-dual methods, Proximal augmented Lagrangian, Operator splitting.
    @INPROCEEDINGS{ozajovCDC22,
    AUTHOR = {I. K. Ozaslan and M. R. Jovanovi\'c},
    TITLE = {Exponential convergence of primal-dual dynamics for multi-block problems under local error bound condition},
    BOOKTITLE = {Proceedings of the 61th IEEE Conference on Decision and Control},
    YEAR = {2022},
    ADDRESS = {Cancun, Mexico},
    PAGES = {7579-7584},
    PDF = {https://viterbi-web.usc.edu/~mihailo/papers/ozajovCDC22.pdf},
    KEYWORDS = {Gradient flow dynamics, Lyapunov functions, Proximal algorithms, Primal-dual gradient flow dynamics, Primal-dual methods, Proximal augmented Lagrangian, Operator splitting} 
    }
    



Back to MJ's Publications




Disclaimer:

This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All person copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.




Last modified: Tue Jan 23 11:32:51 2024
Author: mihailo.


This document was translated from BibTEX by bibtex2html