2023

J. Subramanian, A. Sinha, and A. Mahajan, 鈥淩obustness and sample complexity of model-based MARL for general-sum Markov games,鈥� Dynamic Games and Applications, pp. 56鈥�88, March 2023.

S. Sudhakara, A. Mahajan, A. Nayyar, and Y. Ouyang, 鈥淪calable regret for learning to control network-coupled subsystems with unknown dynamics,鈥� IEEE Transactions on Control of Networked Systems, vol. 10, no. 1, pp. 2鈥�14, March 2023.

M. Afshari and A. Mahajan, 鈥淒ecentralized linear quadratic systems with major and minor agents and non-Gaussian noise,鈥� IEEE Transactions on Automatic Control, vol. 68, no. 8, pp. 4666鈥�4681, Aug 2023

H. Nekoei, A. Badrinaaraayanan, A. Sinha, M. Amini, J. Rajendran, A. Mahajan, and S. Chandar, 鈥淒ealing With Non-stationarity in Decentralized Cooperative Multi-Agent Deep Reinforcement Learning via Multi-Timescale Learning,鈥� Conference on Lifelong Learning Agents, Montreal, Aug 2023.

B. Sayedana, M. Afshari, P.E. Caines, and A. Mahajan, 鈥淎lmost Sure Regret Bounds for Certainty Equivalence Control of Markov Jump Systems,鈥� IEEE Conference of Decision and Control, Singapore, Dec 2023.

A. Sinha and A. Mahajan, 鈥淎symmetric Actor Critic with Approximate Information State,鈥� IEEE Conference on Decision and Control, Singapore, Dec 2023.

B. Bozkurt, A. Mahajan, A. Nayyar, and Y. Ouyang, 鈥淲eighted norm bounds in MDPs with unbounded per-step cost,鈥� IEEE Conference on Decision and Control, Singapore, Dec 2023.

J. Subramanian, A. Sinha, and A. Mahajan, 鈥淩obustness and sample complexity of model-based reinforcement learning for general-sum Markov games,鈥� Dynamics Games and Applications Workshop, Paris, France, Oct 2023.

J. Sumbramanian, A. Kumar, and A. Mahajan, "Mean-field games among teams," PGMODays, Paris, France, Nov 2023.

T. Ni, B. Eysenbach, E. SeyedSalehi , M. Ma, C. Gehring, A. Mahajan, and P.L. Bacon, 鈥淏ridging State and History Representations: Understanding self-predictive RL,鈥� Neurips workshop on Self-Supervised Learning - Theory and Practice, New Orleans, USA, Dec 2023.

G. Patil , A. Mahajan, and D. Precup, 鈥淥n learning history-based policies for controlling Markov decision processes,鈥� ICML Workshop on New Frontiers in Learning, Control, and Dynamical Systems, Hawaii, USA, July 2023.

A. Sinha and A. Mahajan, 鈥淎symmetric Actor Critic with Approximate Information State,鈥� Meeting on Systems and Control Theory, Waterloo, ON, May 2023.

A. Sinha and A. Mahajan, 鈥淎symmetric Actor Critic with Approximate Information State,鈥� Les Cahiers du GERAD, no. G-2023-57, Nov 2023.

B. Sayedana, M. Afshari, P.E. Caines, and A. Mahajan, 鈥淪trong consistency and rate of convergence of switched least squares system identification for autonomous switched Markov jump linear systems,鈥� Les Cahiers du GERAD, no. G-2023-05, Mar 2023.

N. Akbarzadeh and A. Mahajan, 鈥淥n learning Whittle index policy for restless bandits with scalable regret鈥�, Workshop on restless bandits, index policies and applications in reinforcement learning, University of Grenoble-Alpes, Nov 2023.

2022

Jayakumar Subramanian et al. 鈥淎pproximate information state for approximate planning and reinforcement learning in partially observed systems鈥�. In: The Journal of Machine Learning Research 23.1 (2022), pp. 483鈥�565.

Shuang Gao and Aditya Mahajan. 鈥淥ptimal control of network-coupled subsystems: Spectral decomposition and low-dimensional solutions鈥�. In: IEEE Transactions on Control of Network Systems 9.2 (2021), pp. 657鈥�669.

Abhinav Dahiya et al. 鈥淪calable operator allocation for multirobot assistance: A restless bandit approach鈥�. In: IEEE Transactions on Control of Network Systems 9.3 (2022), pp. 1397鈥�1408.

Nima Akbarzadeh and Aditya Mahajan. 鈥淐onditions for indexability of restless bandits and an O (K藛 3) algorithm to compute Whittle index鈥�. In: Annals of Applied Probability 54.4 (2022), pp. 1164鈥�1192.

Anirudha Jitani et al. 鈥淪tructure-aware reinforcement learning for node-overload protection in mobile edge computing鈥�. In: IEEE Transactions on Cognitive Communications and Networking 8.4 (2022), pp. 1881鈥�1897.

Gandharv Patil, Aditya Mahajan, and Doina Precup. 鈥淥n learning history based policies for controlling Markov decision processes鈥�. In: Conference on Reinforcement Learning and Decision Making (RLDM). 2022.

Nima Akbarzadeh and Aditya Mahajan. 鈥淧artially observable restless bandits with restarts: Indexability and computation of Whittle index鈥�. In: 2022 IEEE 61st Conference on Decision and Control (CDC). IEEE. 2022, pp. 4898鈥�4904.

Mukul Gagrani et al. 鈥淎 modified Thompson sampling-based learning algorithm for unknown linear systems鈥�. In: 2022 IEEE 61st Conference on Decision and Control (CDC). IEEE. 2022, pp. 6658鈥�6665.

Hadi Nekoei et al. 鈥淒ealing With Nonstationarity in Decentralized Cooperative Multi-Agent Deep Reinforcement Learning via Multi-Timescale Learning鈥�. In: 2023.

Hadi Nekoei et al. 鈥淪taged independent learning: Towards decentralized cooperative multiagent Reinforcement Learning鈥�. In: ICLR 2022 Workshop on Gamification and Multiagent Solutions. 2022.

2021

M. Afshari and A. Mahajan. 鈥淢ulti-agent estimation and filtering for minimizing team mean-squared error,鈥� IEEE Transactions on Signal Processing, vol. 69, pp. 5206鈥�5221, Aug 2021.

A. Jitani, A. Mahajan, Z. Zhu, H. Abou-zeid, E.T. Fapi, and H. Purmehdi. 鈥淪tructure-aware reinforcement learning for node overload protection in mobile edge computing,鈥� IEEE International Conference on Communications, Montreal, Canada, June 2021.

K. Kaza, A. Mahajan, and J. Le Ny. 鈥淒ecision referrals in human-automation teams,鈥� IEEE Conference on Decision and Control, Austin, TX, Dec 2021.

R. Seraj, A. Mahajan, and J. Le Ny. 鈥淢ean-field approximation for large-population beauty-contest games,鈥� IEEE Conference on Decision and Control, Austin, TX, Dec 2021.

M. Gagrani, S. Sudhakara, A. Mahajan, A. Nayyar, and Y. Ouyang. 鈥淭hompson sampling for linear quadratic mean-field teams,鈥� IEEE Conference on Decision and Control, Austin, TX, Dec 2021. (invited talk)

J. Subramanian, A. Sinha, and A. Mahajan. 鈥淩obustness of Markov perfect equilibrium to model approximations in general-sum dynamic games,鈥� IEEE Indian Control Conference, Mumbai, India, Dec 2021. (invited talk)

N. Akbarzadeh and A. Mahajan. 鈥淢aintenance of a collection of machines under partial observability: Indexability and computation of Whittle index,鈥� Les Cahiers du GERAD, no. G-2021-26, April 2021.

2020

M. Afshari and A. Mahajan, "Optimal local and remote controllers with unreliable uplink channels: An elementary proof." IEEE Transactions on Automatic Control, vol 65. no 8 pp. 3606-3622, Aug 2020

J. Subramanian and A. Mahajan, "Renewal Monte Carlo: Renewal theory based reinforcement learning," IEEE Transactions on Automatic Control, vol. 65, no. 8, pp. 3663-3670. Aug 2020.

B. Sayedana and A. Mahajan, "Counterexamples on the monotonicity of delay optimal strategies for energy harvesting transmitters," IEEE Wireless Communication Letters, vol. 9, no. 7. pp. 1070-1074, Jul 2020

J. Chakravorty and A. Mahajan, "Remote estimation over packet-drop channel with Markovian state," IEEE Transactions on Automatic Control, vol. 65, nno. 5, pp. 2016-2031, May 2020.

B. Sayedana, A. Mahajan, and E. Yeh, "Cross-layer communication over fading channels with adaptive decision feedback," International Symposium on Modeling and Optimization in Mobile, Ad Hov, and Wireless Networks (WiOpt), Jun 2020.

N. Akbarzadeh and A. Mahajan, "Restless bandits, indexability, and computation of Whittle Index," Les Cahiers du GERAD, no G-2020-34, June 2020.

Z. Zhu, H. Abou-zeid, A. Mahajan, and A. Jitani*, "Overload protection for edge cluster using two-tier reinforcement learning models", submitted US patent 4015-11327/P081852WO01

2019

J. Subramanian and A. Mahajan, 鈥淩einforcement learning in stationary mean-field games,鈥� International Conference on Autonomous Agents and Multiagent Systems (AAMAS), Montreal, Canada, 13鈥�17 May, 2019.

J. Subramanian and A. Mahajan, 鈥淎pproximate information state for partially observed systems,鈥� IEEE Conference on Decision and Control, Nice, France, 11鈥�13 Dec, 2019. (invited talk)

N. Akbarzadeh and A. Mahajan, 鈥淩estless bandits with controlled restarts: Indexability and computation of Whittle index,鈥� IEEE Conference on Decision and Control, Nice, France, 11鈥�13 Dec, 2019.

S. Gao and A. Mahajan, 鈥淣etworked control of coupled subsystems: Spectral decomposition and low-dimensional solutions,鈥� IEEE Conference on Decision and Control, Nice, France, 11鈥�13 Dec, 2019.

N. Akbarzadeh and A. Mahajan, 鈥淒ynamic spectrum access under partial observations: A restless bandit approach,鈥� Canadian Workshop on Information Theory (CWIT), Hamilton, Ontario, June 2鈥�5, 2019.

J. Subramanian and A. Mahajan, 鈥淎pproximate information state for partially observed systems,鈥� Conference on Reinforcement Learning and Decision Making (RLDM), Montreal, Canada, 7鈥�10 July, 2019.

J. Subramanian, R. Seraj, and A. Mahajan, 鈥淩einforcement learning for mean-field teams,鈥� Conference on Reinforcement Learning and Decision Making (RLDM), Montreal, Canada, 7鈥�10 July, 2019.

J. Subramanian and A. Mahajan, 鈥淎pproximate information state for partially observed systems,鈥� Neural Information Processing Systems (NeurIPS) Workshop on Optimization Foundations of Machine Learning, Vancouver, Canada, 14 Dec, 2019.

A. Mahajan and J. Subramanian, 鈥淩epresentation Learning via state aggregation: A perspective of control over communication channels,鈥� Neural Information Processing Systems (NeurIPS) Workshop on Information Theory and Machine Learning, Vancouver, Canada, 7 Dec, 2019.

J. Subramanian, R. Seraj, and A. Mahajan, 鈥淩einforcement learning for mean-field teams,鈥� AAMAS Workshop on Adaptive and Learning Agents (ALA), Montreal, Canada, 13鈥�17 May, 2019.

J. Subramanian, A. Kumar, and A. Mahajan, 鈥淢ean-field games between teams,鈥� 11th Workshop on Dynamic Games in Management Science, Montreal, Canada, 24鈥�25 Oct, 2019.

J. Subramanian and A. Mahajan, 鈥淎pproximate dynamic programming and reinforcement learning for partially observed systems,鈥� Montreal AI Symposium, Montreal, Canada, 6 Sep, 2019.

J. Subramanian and A. Mahajan, 鈥淩einforcement learning in stationary mean-field games,鈥� Information Theory and Applications (ITA) Workshop, San Diego, CA, 11鈥�15 Feb, 2019.

J. Subramanian and A. Mahajan, 鈥淩einforcement learning in stationary mean-field games,鈥� Les Cahiers du GERAD, no. G-2019-18, March 2019.

2018

J. Chakravorty and A. Mahajan, 鈥淪ufficient conditions for the value function and optimal strategy to be even and quasi-convex,鈥� IEEE Transactions on Automatic Control, pp. 3858鈥�3864, Nov 2018.

S. Li, A. Khisti, and A. Mahajan, 鈥淚nformation-theoretic privacy for smart metering systems with a rechargeable battery,鈥� IEEE Transactions on Information Theory, pp. 3679鈥�3695, May 2018.

M. Afshari and A. Mahajan, 鈥淭eam optimal decentralized state estimation,鈥� IEEE Conference on Decision and Control, Miami, Florida, Dec 17鈥�19, 2018.

S. Mathew, K.H. Johannson, and A. Mahajan, 鈥淥ptimal sampling of multiple linear processes over a shared medium,鈥� IEEE Conference on Decision and Control, Miami, Florida, Dec 17鈥�19, 2018.

J. Subramanian, A. Mahajan, and A.A. Paranjape, 鈥淥n Controllability of Leader-Follower Dynamics over a Directed Graph,鈥� IEEE Conference on Decision and Control, Miami, Florida, Dec 17鈥�19, 2018.

J. Subramanian and A. Mahajan, 鈥淩enewal Monte Carlo: Renewal theory based reinforcement learning,鈥� IEEE Conference on Decision and Control, Miami, Florida, Dec 17鈥�19, 2018.

M. Afshari and A. Mahajan, 鈥淥ptimal decentralized control of two agent linear system with partial output feedback: certainty equivalence and optimality of linear strategies,鈥� IFAC Workshop on Distributed Estimation and Control in Networked Systems, Groningen, Netherlands, August 27-28, 2018.

J. Subramanian and A. Mahajan, 鈥淎 policy gradient algorithm to compute boundedly rational stationary mean field equilibria,鈥� ICML/IJCAI/AAMAS Workshop on Planning and
Learning (PAL-18), Stockholm, Sweden, July 13鈥�15, 2018

2017

J. Chakravorty and A. Mahajan, 鈥淔undamental limits of remote estimation of autoregressive Markov processes under communication constraints,鈥� IEEE Transactions on Automatic Control, pp. 1109鈥�1124, March 2017.

C. Ma*, A. Mahajan, and B. Meyer, 鈥淢ulti-armed bandits for efficient lifetime estimation in MPSoC design,鈥� Design, Automation and Test in Europe (DATE), Laussane, Switzerland, Mar 23鈥�27, 2017.

J. Chakravorty, J. Subramanian, and A. Mahajan, 鈥淪tochastic approximation based methods for computing the optimal thresholds in remote-state estimation with packet drops,鈥� American Control Conference, Seattle, WA, May 24鈥�26, 2017.

J. Chakravorty and A. Mahajan, 鈥淪tructure of optimal strategies for remote estimation over Gilbert-Elliott channel with feedback,鈥� IEEE International Symposium of Information Theory (ISIT), Aachen, Germany, Jun 25鈥�30, 2017.

M. Afshari and A. Mahajan, 鈥淪tatic teams with common information,鈥� IFAC World Congress, Toulouse, France, Jul 9鈥�14, 2017.

A. Mahajan, 鈥淩emote estimation over control area networks,鈥� IEEE Vehicular Technology Conference (VTC), Networked Vehicles for Intelligent Transportation and Smart Grids (NetV) Workshop, Toronto, Canada, Sep 24鈥�27, 2017.

Y. Liu, A. Khisti, and A. Mahajan, 鈥淥n Privacy in Smart Metering Systems with Periodically Time-Varying Input Distribution,鈥� GlobalSIP Symposium on Control and Information Theoretic Approaches to Security and Privacy, Montreal, Canada, Nov 14鈥�16, 2017.

J.Subramanian, J.Chakravorty, and A.Mahajan,鈥淩enewal theory based reinforcement learning for Markov processes,鈥� Optimization Days, Montreal, QC, May 10鈥�11, 2017.

Mahajan, 鈥淲hen to observe a Markov process,鈥� INFORMS Applied Probability Society Conference, Evanston, IL, July 10鈥�12, 2017.

M. Afshari and A. Mahajan, 鈥淒ecentralized Kalman Filtering,鈥� Fields Institute Workshop on Stochastic Processes and their Applications, Carleton University, Ottawa, ON, Aug 9鈥�11, 2017.

J. Subramanian and A. Mahajan, 鈥淎 new policy based RL algorithm with reduced bias and variance,鈥� Montreal AI Symposium, Montreal, QC, Sep 26, 2017.

M. Afshari and A. Mahajan, 鈥淭eam optimal decentralized filtering with coupled cost,鈥� Ninth Workshop on Dynamic Games in Management Science, Montreal, QC, Oct 12鈥�13, 2017.

M. Afshari and A. Mahajan, 鈥淪tatic teams with common information,鈥� Les Cahiers du GERAD, no. G-2017-29, April 2017.