Emam Hossain

iHARP Research Assistant | UMBC Ph.D. Student

Email: emamh1@umbc.edu
See me at Google Scholar Citation Page, LinkedIn

Research Interests

  • Causality
  • Machine Learning
  • Deep Learning
  • Climate Science

Successfully defended the Ph.D. proposal in June 2025

Short Biography

Emam Hossain is a Ph.D. candidate in the Department of Information Systems at the University of Maryland, Baltimore County (UMBC), where he also earned his M.S. in Information Systems in 2023. He previously received his Bachelor’s and Master’s degrees in Computer Science from the University of Chittagong, Bangladesh. His research interests lie at the intersection of machine learning, deep learning, and causality, with a strong emphasis on their application in scientific domains such as climate science. Emam has authored several peer-reviewed publications in top-tier journals and conferences, contributing to methodological advancements in time-series modeling and causal analysis. At UMBC, he is an active member of the iHARP research team, where he collaborates on interdisciplinary efforts to address climate-related challenges through data-driven approaches.

Research Summary

Emam’s research centers on advancing machine learning methods to uncover and model causal mechanisms in complex environmental systems, with a particular focus on the Arctic and Greenland Ice Sheet. A core aspect of his work involves supraglacial lakes—meltwater bodies that form on the surface of glaciers during the summer melt season. These lakes are not merely indicators of surface melt but active agents in glacier dynamics, as their drainage and expansion influence meltwater runoff pathways and can trigger rapid ice mass loss. Emam develops causally-informed time-series modeling frameworks that go beyond correlation-based predictions to identify and quantify the direct drivers of such phenomena. His work introduces novel approaches for learning stable, interpretable representations of environmental processes by embedding causal structure into temporal models. These contributions aim to improve the robustness and generalizability of climate forecasts under non-stationary conditions, ultimately supporting more reliable scientific understanding and climate adaptation strategies.

Publications
  • Emam Hossain, Muhammad Hasan Ferdous, Jianwu Wang, Aneesh Subramanian, and Md Osman Gani (2025). Correlation to Causation: A Causal Deep Learning Framework for Arctic Sea Ice Prediction. In 2025 IEEE International Conference on Pervasive Computing and Communications (PerCom). IEEE.
  • Devon Dunmire, Aneesh C. Subramanian, Emam Hossain, Md Osman Gani, Alison F. Banwell, Hammad Younas, and Brendan Myers. 2025. Greenland Ice Sheet Wide Supraglacial Lake Evolution and Dynamics: Insights From the 2018 and 2019 Melt Seasons. Earth and Space Science 12, 2 (February 2025), e2024EA003793. https://doi.org/10.1029/2024EA003793
  • Emam Hossain, Md Osman Gani, Devon Dunmire, Aneesh Subramanian, and Hammad Younas (2024). Time Series Classification of Supraglacial Lakes Evolution over Greenland Ice Sheet. In 2024 International Conference on Machine Learning and Applications (ICMLA). IEEE.Miami, Florida (https://arxiv.org/abs/2410.05638)
  • Aneesh Subramanian, Devon Dunmire, Emam Hossain, Md Osman Gani, Alison Banwell, and Brendan Myers. 2024. The fate of Greenland Ice Sheet supraglacial lakes in a warm and cool year. https://doi.org/10.5194/egusphere-egu24-20925
  • Emam Hossain, Sahara Ali, Yiyi Huang, Nicole J Shchlegel, Jianwu Wang, Aneesh C. Subramanian, & Md Osman Gani. “Incorporating Causality with Deep Learning in Predicting Short-term and Seasonal Sea Ice”. Abstract in 23rd Conference on Artificial Intelligence for Environmental Science, 104th AMS Annual Meeting, 2024, November 2023. [Abstract]
  • Uzma Hasan, Emam Hossain, and Md Osman Gani. “A Survey on Causal Discovery Methods for I.I.D. and Time Series Data,” 2023. https://doi.org/10.48550/ARXIV.2303.15027. [Journal]
Proposal Defense

Successfully Defended On: June 5, 2025

Abstract

Causal reasoning is fundamental to understanding dynamic processes in complex scientific systems such as climate, glaciology, and remote sensing. While modern deep learning models have demonstrated strong performance on sequence modeling tasks, they often rely on purely correlational representations, limiting their generalization and interpretability under distribution shifts or partial observability. This dissertation presents a principled progression of methods that integrate causal discovery and representation learning for modeling dynamic systems. First, we reinterpret Reconstructed Phase Space (RPS) modeling as an unsupervised representation learning approach for satellite-derived time series, enabling interpretable classification of supraglacial lake evolution via Gaussian Mixture Models. We then extend this with a causally-informed temporal modeling framework using PCMCI+ to identify stable, regionally meaningful predictors, which improve generalization and robustness in sequence classification models. However, both methods operate in the space of observed variables and do not recover unobserved mechanisms. To address this limitation, we propose a novel Causal Representation Learning (CRL) framework for dynamic systems that employs variational autoencoders with interventional regularization to learn low-dimensional latent variables aligned with underlying causal factors. This framework is designed to support do-operations and counterfactual reasoning in latent space, and will be validated on benchmark datasets with known causal graphs (Pendulum, Flow, CelebA), before being applied to supraglacial lake dynamics as a real-world use case. Collectively, this work aims to advance the boundary of causal modeling in temporal systems by bridging the gap between statistical representation learning and structural causal inference, offering scalable and interpretable tools for scientific discovery in dynamic, high-dimensional environments.

Committee

  • Dr. Md Osman Gani – Advisor and Committee Chair (UMBC)
  • Dr. Vandana Janeja – Committee Member (UMBC)
  • Dr. James Foulds – Committee Member (UMBC)
  • Dr. Aneesh Subramanian – Committee Member (UC Boulder)
  • Dr. Devon Dunmire – Committee Member (KU Leuven)

Last updated 16 June 2025