Avinash Mohan

Assistant Professor, Computer Science

With more than a decade of experience in reinforcement learning, machine learning, and computer networking, Dr. Mohan’s research focuses on statistical learning, including reinforcement learning, online learning (multi-armed bandit problems), game theory, and convex optimization. He also researches computer networks and communication, including queueing theory, stochastic processes, information theory, resource allocation and network control. He was a postdoctoral fellow at Boston University (where he was awarded the Distinguished Educator-Scholar Fellowship for 2022–2023), the University of Michigan, and the Technion–Israel Institute of Technology, where he was awarded the PBC Fellowship. He is recipient of the Best Presentation Award (MAC layer track) at the IEEE International Conference on Computer Communications (Infocom) 2018.

Research Interests

Courses

Scholarly Works

Preprints and Techreports

F. Fayza, C. Demirkiran, C. Chen, C. Liu, A. Mohan, H. Errahmouni, S. Yun, M. Imani, D. Zhang, D. Bunandar, and A. Joshi. “Towards Efficient Hyperdimensional Computing Using Photonics” (2023). https://arxiv.org/pdf/2311.17801

Zaki, M., Mohan, A., Gopalan, A., and Mannor, S. “Actor-Critic based Improper Reinforcement Learning” (2022).

Zaki, M., Mohan, A., and Gopalan, A. “Improved Pure Exploration in Linear Bandits with No-Regret Learning” (2021). https://doi.org/10.13140/RG.2.2.35458.20162

Mohan, A., Vatsa, S., Kumar, A., and Chattopadhyay, A. “Decentralized, Hybrid MAC Design with Reduced State Information Exchange for Low-Delay IoT Applications” (2021). https://arxiv.org/pdf/2105.11213

Zaki, M., Mohan, A., Gopalan, A., and Mannor, S. “Improper Learning with Gradient-based Policy Optimization” (2021). https://arxiv.org/abs/2102.08201

Mohan, A., Mannor, S., and Kizilkale, A. C. “On the Volatility of Optimal Control Policies and the Capacity of a Class of Linear Quadratic Regulators” (2020). https://arxiv.org/pdf/2002.06808

Zaki, M., Mohan, A., and Gopalan, A. “Explicit Best Arm Identification in Linear Bandits Using No-Regret Learners” (2020). https://arxiv.org/pdf/2006.07562

Mohan, A., Gopalan, A., and Kumar, A. “Throughput Optimal Decentralized Scheduling with Single-bit State Feedback for a Class of Queueing Systems” (2020). https://arxiv.org/pdf/2002.08141

Zaki, M., Mohan, A., and Gopalan, A. “Towards Optimal and Efficient Best Arm Identification in Linear Bandits” (2019). https://arxiv.org/abs/1911.01695

Conference Papers

Zaki, M., Mohan, A., Gopalan, A., and Mannor, S. “Actor Critic Based Improper Reinforcement Learning.” Proceedings of the 39th International Conference on Machine Learning (ICML), in Proceedings of Machine Learning Research vol. 162 (July 17–23, 2022): 25867–25919. https://proceedings.mlr.press/v162/zaki22a/zaki22a.pdf

Zaki, M., Mohan, A., and Gopalan, A. “Explicit Best Arm Identification in Linear Bandits Using No-Regret Learners.” International Conference on Artificial Intelligence (IJCAI-ECAI), 2022. https://arxiv.org/pdf/2006.07562

Zaki, M., Mohan, A., Gopalan, A., and Mannor, S. “Actor Critic Based Improper Reinforcement Learning.” The Multi-disciplinary Conference on Reinforcement Learning and Decision Making (RLDM), 2022.

Zaki, M., Mohan, A. and Gopalan, A., and Mannor, S. “Better than the Best: Gradient-based Improper Reinforcement Learning for Network Scheduling.” Reinforcement Learning in Networks and Queues Workshop, ACM SIGMETRICS, 2021. https://arxiv.org/pdf/2105.00210

Mohan, A., Mannor, S., and Kizilkale, A. C. “On the Volatility of Optimal Control Policies of a Class of Linear Quadratic Regulators.” The 2021 American Control Conference (ACC), 2021. https://arxiv.org/pdf/2002.06808

Vatsa, S., Mohan, A., and Kumar, A. “Implementing QZMAC (a Decentralized Delay Optimal MAC) over 6TiSCH under the Contiki OS in an IEEE 802.15.4 Network.” The Demos and Exhibits workshop at the International Conference on COMmunication Systems & NETworkS (COMSNETS), 2021. Runner Up, Best Research Demo Award. https://arxiv.org/pdf/2012.02955

Zaki, M., Mohan, A., and Gopalan, A. “Employing No Regret Learners for Pure Exploration in Linear Bandits.” Optimization for Machine Learning Workshop at the 34th Conference on Neural Information Processing Systems (NeurIPS), 2020. Selected for Spotlight presentation.

Zaki, M., Mohan, A., and Gopalan, A. “Towards Optimal and Efficient Best Arm Identification in Linear Bandits.” CausalML Workshop at the 33rd Conference on Neural Information Processing Systems (NeurIPS), Vancouver, Canada, 2019.

Mohan, A., Gopalan, A., and Kumar, A. “Reduced-State, Optimal Medium Access Control for Wireless Data Collection Networks.” IEEE International Conference on Computer Communications (INFOCOM), 2018. Best Presentation Award.

Mohan, A., Chattopadhyay, A., and Kumar, A. “Hybrid MAC Protocols for Low-delay Scheduling.” IEEE International Conference on Mobile Ad hoc and Sensor Systems (IEE MASS), 2016. IEEE CompSoc Student Travel Award.

Mohan, A., and Hari, K. V. S. “Low complexity adaptation for channel shortening equalizers.” IEEE 54th International Midwest Symposium on Circuits and Systems (MWSCAS), 2011.

Journal Papers

Mohan, A., Chattopadhyay, A., Vatsa, S. V., and Kumar, A. “A Low-Delay MAC for IoT Applications: Decentralized Optimal Scheduling of Queues Without Explicit State Information Sharing.” IEEE Transactions on Control of Network Systems (2024): 1–12. https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=10584286

Mohan, A., Gopalan, A., and Kumar, A. “Reduced-State, Optimal Scheduling for Decentralized Medium Access Control of a Class of Wireless Networks.” IEEE/ACM Transactions on Networking 28, no. 3 (2020): 1017–1032. https://ieeexplore.ieee.org/abstract/document/9078869

Mohan, A., and Hari, K. V. S. “Low complexity adaptation for SISO channel shortening equalizers.” AEU-International Journal of Electronics and Communications 66, no. 8 (2012): 600–604. https://doi.org/10.1016/j.aeue.2012.03.011

View all profiles