Poushali Sengupta

I am currently a researcher at the RESPiRE project (Responsible Explainable Machine Learning for Sleep-related Respiratory Disorders) within the DKM group at the Department of Informatics and the dScience Center, University of Oslo (UiO), Norway, where I work with Professor Thomas Plagemann on trustworthy and explainable machine learning methods for healthcare-related applications. I am also in the final phase of my PhD research in the ND (Network and Distributed Systems) group at UiO, under the supervision of Professors Sabita Maharjan, Yan Zhang, and Frank Eliassen. 

My research lies at the intersection of explainable and trustworthy AI, statistical learning, reliability analysis, and privacy-aware intelligent systems. I have worked extensively on correlation- and dependence-aware explainability methods, the robustness and reliability of AI explanations, conformal prediction, uncertainty quantification, and privacy-preserving AI frameworks for critical infrastructure, healthcare, communication systems, and energy applications.

I have authored and co-authored multiple journal articles, conference papers, book chapters, and ongoing works under review in the areas of trustworthy AI, explainable machine learning, privacy, and reliable AI systems. My recent research includes ExCIR, MCIR, Kernel-MCIR, reliability-aware explainability frameworks, and privacy-preserving hierarchical AI systems.

Alongside my research, I have been involved in teaching activities at the University of Oslo as a teaching assistant for the IN5410/IN9410 Energy Informatics course at the master’s and PhD levels. My teaching activities include supporting students with assignments, machine learning projects, and energy forecasting applications. I have also supervised bachelor’s and master’s thesis projects related to explainable and trustworthy AI. In addition, I completed the 200-hour Basic Competence in University Pedagogy program at UiO.

I actively collaborate across interdisciplinary and international research environments, including collaborations with Simula Research Laboratory, SINTEF, Aalborg University, OsloMet, and other Nordic institutions. I was also an active student council member of NORA (Norwegian Artificial Intelligence Research Consortium), contributing to interdisciplinary AI research and to the development of the academic community across Norway.

My broader research vision is to develop AI systems that are not only accurate and scalable but also transparent, reliable, privacy-aware, human-centered, and societally responsible. 

Outside of research, I enjoy painting, traveling, and exploring different cultures and perspectives. I also love discovering new places, random cafés, and small everyday adventures, which probably explains why I enjoy interdisciplinary research so much, too ;)

PUBLICATIONS

1. 1. Sengupta, P., Paul, S., and Mishra, S. (2021). “Learning with Differential Privacy.” In Handbook of Research on Cyber Crime and Information Privacy, IGI Global, pp. 372–395.
2. 2. Sengupta, P., Zhang, Y., Maharjan, S., and Eliassen, F. (2023). “Balancing Explainability-Accuracy of Complex Models.” arXiv preprint arXiv:2305.14098.
3. 3. Sengupta, P., Maharjan, S., Eliassen, F., and Zhang, Y. (2025). “Privacy-Utility-Fairness: A Balanced Approach to Vehicular-Traffic Management System.” Proceedings of the IEEE 101st Vehicular Technology Conference (VTC Spring), pp. 1–6.
4. 4. Sengupta, P., Paul, S., and Mishra, S. (2020). “BUDS: Balancing Utility and Differential Privacy by Shuffling.” IEEE Proceedings of the 11th International Conference on Computing, Communication and Networking Technologies (ICCCNT), pp. 1–7.
5. 5. Sengupta, P., Paul, S., and Mishra, S. (2023). “BUDS+: Better Privacy with Converger and Noisy Shuffling.” ACM Digital Threats: Research and Practice, 4(2), pp. 1–23.
6. 6. Sengupta, P. and Mishra, S. (2021). “Fairly Private Through Group Tagging and Relation Impact.” International Conference on Modeling Decisions for Artificial Intelligence, Springer, pp. 259–272.
7. 7. Duguma, D. G., Zhang, J., Aboutalebi, M., Zhang, S., Banet, C., Bjørkli, C., Baramashetru, C., Eliassen, F., Zhang, H., Muringani, J., et al. (2023). “Privacy-Preserving Transactive Energy Systems: Key Topics and Open Research Challenges.” arXiv preprint arXiv:2312.11564.
8. 8. Sengupta, P., Raikwar, M., Zhang, Y., Maharjan, S., and Eliassen, F. (2025). “Context-Aware Adaptive Post-Quantum Framework for Autonomous Vehicular Networks.” Accepted at FutureG Workshop at NDSS 2026.
9. 9. Sengupta, P. (2026). “Dependence-Aware Context Selection for Trustworthy Autoregressive Language Models.” arXiv preprint arXiv:2602.01378. Under review.
10. 10. Sengupta, P., Zhang, Y., and Maharjan, S. (2025). “HXAI: A Privacy-Preserving Hierarchical Explainable AI Framework for Smart Energy Systems.” Under review at IEEE Transactions on Emerging Topics in Computational Intelligence (TETCI).
11. 11. Sengupta, P., Khadka, R., Maharjan, S., Eliassen, F., Lind, P. G., Yazidi, A., and Zhang, Y. (2025). “ExCIR: An Approach Towards Robust Explainability of Complex Models Through Correlation Impact Ratio.” Under review at IEEE Transactions on Artificial Intelligence.
12. 12. Sengupta, P., Maharjan, S., Eliassen, F., Zhang, Y., and Pandey, S. R. (2025). “MCIR: Mutual-Information Ratio for Correlation-Aware Feature Attribution.” Manuscript under submission to Transactions on Machine Learning Research (TMLR).
13. 13. Sengupta, P., Maharjan, S., Eliassen, F., and Zhang, Y. (2026). “Kernel-MCIR: A Kernelized Correlation-Aware Attribution Framework for High-Dimensional Feature Redundancy.” Under review at NeurIPS.
14. 14. Sengupta, P., Maharjan, S., Eliassen, F., Vinel, A., and Zhang, Y. (2026). “EC-NCS: Explanation-Conditioned Conformal Prediction for Reliable Machine Learning in Communication Networks.” Manuscript under review at Globecom 2026.
15. 15. Sengupta, P., Zhang, Y., Eliassen, F., and Maharjan, S. (2026). “Correlation-Aware Feature Attribution-Based Explainable AI.” Accepted in Proceedings of the International Conference on Advances in Artificial Intelligence and Machine Learning (AAIML), 2026, Japan, Tokyo.
16. 16. Khadka, R., Sengupta, P., Lind, P. G., and Yazidi, A. (2024). “Quantization of Vision Transformer-Based Model for Real-Time EEG Classification.” Nordic e-Infrastructure Collaboration Conference, Springer Nature Switzerland, pp. 17–27.
17. 17. Paul, S., Sengupta, P., and Mishra, S. (2020). “FLAPS: Federated Learning and Privately Scaling.” Proceedings of the IEEE 17th International Conference on Mobile Ad Hoc and Sensor Systems (MASS), pp. 13–19.
18. 18. Sengupta, P., Maharjan, S., Eliassen, F., Pandey, S. R., and Zhang, Y. (2026). “Reliable Explanations or Random Noise? A Reliability Metric for XAI.” Under review at NeurIPS.
19. 19. Sengupta, P., Maharjan, S., Eliassen, F., and Zhang, Y. (2026). “X-NegoBox: An Explainable Privacy-Budget Negotiation Framework for Secure Peer-to-Peer Energy Data Exchange.” Accepted at IEEE International Conference on Computer Communications and Networks (ICCCN), 2026, Honolulu, Hawaii.