Haoran Zhang
PhD Student
Electrical Engineering and Computer Science
Massachusetts Institute of Technology
Email: haoranz [at] mit [dot] edu
GitHub
Google Scholar
About
I am a fourth-year PhD student at
MIT LIDS, where I am advised by
Dr. Marzyeh Ghassemi. My research generally focuses on improving the reliability of machine learning models in healthcare. Broadly, I am interested in the following topics:
- Algorithmic fairness beyond group fairness: What does it mean for a model to be fair in the clinical setting, and how can we achieve these definitions algorithmically?
- Machine learning under real-world distribution shifts: Machine learning models frequently experience performance drops in new domains, specific subpopulations, and over time. How can we attribute the root cause of such degradations, and construct robust models which generalize?
- Machine learning with noisy labels: Labels in real-world healthcare datasets frequently originate from humans or error-prone automated processes. How can we effectively identify noisy samples, or construct methods to learn fair models under such noise?
I previously completed my MSc at the
University of Toronto under the co-supervision of
Dr. Marzyeh Ghassemi and
Dr. Quaid Morris. I am currently at Stanford University for the summer visiting
Dr. Sanmi Koyejo and
Dr. Nigam Shah.
(* denotes equal contribution)
Preprints
MedPAIR: Measuring Physicians and AI Relevance Alignment in Medical Question Answering
Yuexing Hao, Kumail Alhamoud, Hyewon Jeong, Haoran Zhang, Isha Puri, Philip Torr, Mike Schaekermann, Ariel D. Stern, Marzyeh Ghassemi
[Paper] [Code]
2025
LEMoN: Label Error Detection using Multimodal Neighbors
Haoran Zhang*, Aparna Balagopalan*, Nassim Oufattole, Hyewon Jeong, Yan Wu, Jiacheng Zhu, Marzyeh Ghassemi
International Conference on Machine Learning (ICML) 2025
[Paper] [Code]
2024
BendVLM: Test-Time Debiasing of Vision-Language Embeddings
Walter Gerych, Haoran Zhang, Kimia Hamidieh, Eileen Pan, Maanas Sharma, Thomas Hartvigsen, Marzyeh Ghassemi
Conference on Neural Information Processing Systems (NeurIPS) 2024
[Paper] [Code]
A Closer Look at AUROC and AUPRC under Class Imbalance
Matthew B. A. McDermott, Haoran Zhang, Lasse Hyldig Hansen, Giovanni Angelotti, Jack Gallifant
Conference on Neural Information Processing Systems (NeurIPS) 2024
[Paper]
Identifying Implicit Social Biases in Vision-Language Models
Kimia Hamidieh, Haoran Zhang, Walter Gerych, Thomas Hartvigsen, Marzyeh Ghassemi
Seventh AAAI/ACM Conference on AI, Ethics, and Society (AIES-24)
[Paper]
DOSSIER: Fact Checking in Electronic Health Records While Preserving Patient Privacy
Haoran Zhang, Supriya Nagesh, Milind Shyani, Nina Mishra
Machine Learning for Healthcare (MLHC) 2024
[Paper] [Code]
The Limits of Fair Medical Imaging AI in Real World Generalization
Yuzhe Yang*, Haoran Zhang*, Judy W. Gichoya, Dina Katabi, Marzyeh Ghassemi
Nature Medicine
[Paper] [Code] [News]
Views Can Be Deceiving: Improved SSL Through Feature Space Augmentation
Kimia Hamidieh, Haoran Zhang, Swami Sankaranarayanan, Marzyeh Ghassemi
International Conference on Learning Representations (ICLR) 2024 (Spotlight)
[Paper]
2023
"Why did the Model Fail?": Attributing Model Performance Changes to Distribution Shifts
Haoran Zhang*, Harvineet Singh*, Marzyeh Ghassemi, Shalmali Joshi
International Conference on Machine Learning (ICML) 2023
[Paper] [Code]
Change is Hard: A Closer Look at Subpopulation Shift
Yuzhe Yang*, Haoran Zhang*, Dina Katabi, Marzyeh Ghassemi
International Conference on Machine Learning (ICML) 2023
[Paper] [Code]
Algorithmic Fairness in Chest X-ray Diagnosis: A Case Study
Haoran Zhang, Thomas Hartvigsen, Marzyeh Ghassemi
MIT Case Studies in Social and Ethical Responsibilities of Computing (SERC)
[Paper]
2022
PAN-cODE: COVID-19 Forecasting Using Conditional Latent ODEs
Ruian Shi, Haoran Zhang, Quaid Morris
Journal of the American Medical Informatics Association (JAMIA)
[Paper] [Code]
The Road to Explainability is Paved with Bias: Measuring the Fairness of Explanations
Aparna Balagopalan, Haoran Zhang, Kimia Hamidieh, Thomas Hartvigsen, Frank Rudzicz, Marzyeh Ghassemi
ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT) 2022
[Paper] [Code] [News]
Improving the Fairness of Chest X-ray Classifiers
Haoran Zhang, Natalie Dullerud, Karsten Roth, Lauren Oakden-Rayner, Stephen Robert Pfohl, Marzyeh Ghassemi
Conference on Health, Inference, and Learning (CHIL) 2022
[Paper] [Code]
Reading Race: AI Recognises Patient's Racial Identity In Medical Images
Imon Banerjee, Ananth Reddy Bhimireddy, [...], Zachary Zaiman, Haoran Zhang, Judy W Gichoya
The Lancet Digital Health
[Paper] [Code] [News]
A Comparison of Approaches to Improve Worst-Case Predictive Model Performance over Patient Subpopulations
Stephen R. Pfohl, Haoran Zhang, Yizhe Xu, Agata Foryciarz, Marzyeh Ghassemi, Nigam H. Shah
Scientific Reports
[Paper] [Code]
2021
Underdiagnosis Bias of Artficial Intelligence Algorithms Applied to Chest Radiographs in Under-served Patient Populations
Laleh Seyyed-Kalantari, Haoran Zhang, Matthew McDermott, Irene Y Chen, Marzyeh Ghassemi
Nature Medicine
[Paper]
Learning Optimal Predictive Checklists
Haoran Zhang, Quaid Morris, Berk Ustun, Marzyeh Ghassemi
Conference on Neural Information Processing Systems (NeurIPS) 2021
[Paper] [Code] [Poster]
Pulling Up by the Causal Bootstraps: Causal Data Augmentation for Pre-training Debiasing
Sindhu C.M. Gowda, Shalmali Joshi, Haoran Zhang, Marzyeh Ghassemi
Conference on Information and Knowledge Management (CIKM) 2021
[Paper] [Code] [Poster]
An Empirical Framework for Domain Generalization in Clinical Settings
Haoran Zhang, Natalie Dullerud, Laleh Seyyed-Kalantari, Quaid Morris, Shalmali Joshi, Marzyeh Ghassemi
ACM Conference on Health, Inference, and Learning (CHIL) 2021
[Paper] [Code] [Poster]
2020
An Empirical Study of Representation Learning for Reinforcement Learning in Healthcare
Taylor Killian, Haoran Zhang, Jayakumar Subramanian, Mehdi Fatemi, Marzyeh Ghassemi
Machine Learning for Health Workshop at NeurIPS 2020 (Proceedings Track)
[Paper] [Code] [Poster]
Self-Supervised Contrastive Learning of Protein Representations By Mutual Information Maximization
Amy Lu, Haoran Zhang, Marzyeh Ghassemi, Alan Moses
bioRxiv Preprint
[Paper] [Code]
Hurtful Words: Quantifying Biases in Clinical Contextual Word Embeddings
Haoran Zhang*, Amy Lu*, Mohamed Abdalla, Matthew McDermott, Marzyeh Ghassemi
ACM Conference on Health, Inference, and Learning (CHIL) 2020
[Paper] [Code] [Poster]
Identifying Transitional High Cost Users from Unstructured Patient Profiles Written by Primary Care Physicians
Haoran Zhang, Elisa Candido, Andrew S. Wilton, Raquel Duchen, Liisa Jaakkimainen, Walter Wodchis, Quaid Morris
Pacific Symposium on Biocomputing (PSB) 2020
[Paper] [Poster]
