Susobhan Ghosh

I am currently a fifth year PhD student in the StatRL research group at Harvard University, working with Prof. Susan Murphy. I am currently focusing on designing Bayesian Reinforcement Learning algorithms for Mobile Health interventions through clinical trials. Recently, we developed and deployed the reBandit algorithm for the MiWaves clinical trial (Mar to May 2024), aimed at reducing cannabis use among emerging adults (ages 18-25).

My past work has dealt with problems in the domain of Multi-Agent Systems, Game Theory & Mechanism Design, and Machine Learning, and I have experience applying them to mobile health settings, computational sustainability problems, social problems like security and planning, and adversarial settings.

Publications

  • “It felt more real”: Investigating the User Experience of the MiWaves Personalizing JITAI Pilot Study
    Under Review [PDF]
    Susobhan Ghosh, Yongyi Guo, Pei-Yao Hung, Lara Coughlin, Erin Bonar, Inbal Nahum-Shani, Maureen Walton, Susan Murphy

  • Effective Monitoring of Online Decision-Making Algorithms in Digital Intervention Implementation
    Under Review [PDF]
    Susobhan Ghosh*, Anna Trella*, Erin Bonar, Lara Coughlin, Finale Doshi-Velez, Yongyi Guo, Pei-Yao Hung, Inbal Nahum-Shani, Vivek Shetty, Maureen Walton, Iris Yan, Kelly Zhang, Susan A Murphy

  • A mobile health intervention for emerging adults with regular cannabis use: A micro-randomized pilot trial design protocol
    Contemporary Clinical Trials [PDF]
    Lara Coughlin, Maya Campbell, Tiffany Wheeler, Chavez Rodriguez, Autumn Florimbio, Susobhan Ghosh, Yongyi Guo, Pei-Yao Hung, Mark Newman, Huijie Pan, Kelly Zhang, Lauren Zimmermann, Erin Bonar, Maureen Walton, Susan Murphy, Inbal Nahum-Shani

  • ReBandit: Random Effects Based Online RL Algorithm for Reducing Cannabis Use
    IJCAI 2024 [PDF]
    Susobhan Ghosh, Yongyi Guo, Pei-Yao Hung, Lara Coughlin, Erin Bonar, Inbal Nahum-Shani, Maureen Walton, Susan Murphy

  • Did we personalize? assessing personalization by an online reinforcement learning algorithm using resampling
    Machine Learning Journal [PDF]
    Susobhan Ghosh*, Raphael Kim*, Prasidh Chhabria, Raaz Dwivedi, Predrag Klasnja, Peng Liao, Kelly Zhang, Susan Murphy

  • VidyutVanika: AI-Based Autonomous Broker for Smart Grids: From Theory to Practice
    In the proceedings of Energy Sustainability through Retail Electricity Markets: The Power Trading Agent Competition (Power TAC) Experience [PDF]
    Sanjay Chandlekar, Bala Pedasingu, Susobhan Ghosh, Easwar Subramanian, Sanjay Bhat, Praveen Paruchuri, Sujit Gujar

  • Fairness for Workers Who Pull the Arms: An Index Based Policy for Allocation of Restless Bandit Tasks
    AAMAS 2023 [PDF]
    Arpita Biswas, Jackson Killian, Paula Diaz, Susobhan Ghosh, Milind Tambe

  • ArchGym: An Open-Source Gymnasium for Machine Learning Assisted Architecture Design
    ISCA 2023 [PDF]
    Srivatsan Krishnan, Amir Yazdanbakhsh, Shvetank Prakash, Jason Jabbour, Ikechukwu Uchendu, Susobhan Ghosh, Behzad Boroujerdian, Daniel Richins, Devashree Tripathy, Aleksandra Faust, Vijay Reddi

  • Using Public Data to Predict Demand for Mobile Health Clinics
    AAAI 2022 [PDF]
    Haipeng Chen, Susobhan Ghosh, Gregory Fan, Nikhil Behari, Arpita Biswas, Mollie Williams, Nancy Oriol, Milind Tambe

  • Facilitating Human-Wildlife Cohabitation through Conflict Prediction
    AAAI 2022 [PDF]
    Susobhan Ghosh, Pradeep Varakantham, Aniket Bhatkhande, Tamanna Ahmad, Anish Andheria, Wenjun Li, Aparna Taneja, Divy Thakkar, Milind Tambe

  • Efficient Algorithms for Finite Horizon and Streaming Restless Multi-Armed Bandit Problems
    AAMAS 2022 [PDF]
    Aditya Mate, Arpita Biswas, Christoph Siebenbrunner, Susobhan Ghosh, Milind Tambe

  • Bidding in Smart Grid PDAs: Theory, Analysis and Strategy
    AAAI 2020 [PDF]
    Susobhan Ghosh, Sujit Gujar, Praveen Paruchuri, Easwar Subramanian, Sanjay Bhat

  • VidyutVanika: A Reinforcement Learning Based Broker Agent for a Power Trading Competition
    AAAI 2019 [PDF]
    Susobhan Ghosh, Easwar Subramanian, Sanjay Bhat, Sujit Gujar, Praveen Paruchuri


Talks

  • Deploying RL Algorithms for Digital Interventions (ENAR 2025)
  • reBandit: Personalizing Treatment Delivery for Reducing Cannabis use (INFORMS Annual Meeting 2024)
  • reBandit: Random Effects based Online RL algorithm for Reducing Cannabis Use (IJCAI 2024)
  • MiWaves: AI-driven Digital Health Interventions to help reduce cannabis use (SAA 2024)
  • RL in real life: mixed effects models and reBandit (Invited Lecture, STAT 234 @ Harvard taught by Prof. Susan Murphy)
  • MiWaves: Using AI to help reduce cannabis use among emerging adults (d3C Think Tank Jan 2024)
  • MiWaves: developing a 2nd Generation JITAI to reduce cannabis use among emerging adults (d3C Think Tank May 2023)
  • MiWaves (d3C Think Tank Oct 2022)
  • VidyutVanika: A Reinforcement Learning Based Broker Agent for a Power Trading Competition (AAAI 2019)
  • Review of Sidebar and it’s functionality (LibreOffice Conference 2016)

Posters

  • reBandit: Random Effects based Online RL algorithm for Reducing Cannabis Use (IJCAI 2024)
  • MiWaves: AI-driven Digital Health Interventions to help reduce cannabis use (SCT 2024)
  • MiWaves: AI-based mHealth intervention to reduce cannabis use among emerging adults (mDOT Annual Meeting)
  • MiWaves: AI-based mobile health intervention to reduce cannabis use amongst emerging adults (Dartmouth Digital Mental Health & AI Symposium)
  • VidyutVanika: A Reinforcement Learning Based Broker Agent for a Power Trading Competition (AAAI 2019)

Contact

Work Email: susobhan_ghosh@g.harvard.edu
Personal Email: susobhang70@gmail.com