Haoran Zhang

Hi! I'm currently a second-year MS student in Electrical and Computer Engineering Department at Carnegie Mellon University (CMU). Now I am working with Prof. Carlee Joe-Wong on federated learning. I am also working with Dr. Marie Siew and Prof. Rachid El-Azouzi in the current research project.

Previously, I obtained my bachelor's degree in Automation from Huazhong University of Science and Technology. During the undergrad study, I worked on AI for healthcare with Prof. Hao Chen at HKUST, and Dr. Zhongliang Jiang at TUM.

I will graduate from CMU in May 2025, and after that, begin my PhD study at University of Texas at Austin this fall.

Email  /  CV  /  Scholar  /  Github /  LinkedIn

Research

My current project is about multi-model federated learning (MMFL): local clients can handle multiple unrelated FL models concurrently. For example, a smartphone user can generate data for both keyboard prediction and speech recognition models. Instead of building two single-model FL systems, we build one MMFL system and solve system-level optimization problems to make it more efficient.

My research at CMU: Federated Learning and Optimization

This work extends our preliminary result (Best Poster Award at ICDCS), which focused on minimizing update variance to stabilize training in MMFL. While reducing variance is beneficial, our new analysis shows that it is only part of the picture. We revisit the convergence behavior of MMFL and reveal that optimizing solely for variance ignores other critical factors that significantly impact convergence—especially under heterogeneous system constraints. In response, we propose new solutions that not only achieve better performance but also come at a much lower cost, paving the way for practical and scalable MMFL deployments.

This paper explores an MMFL scenario where each client is restricted to training only one model per global round due to limited computational resources. We propose a client sampling and task assignment method that theoretically minimizes the variance of global updates. The method achieves a 30% higher accuracy than the baseline.

In the real-world scenario, clients could contribute to multiple federated learning models' training at the same time. For example, a company may update their keyboard prediction model, speech recognition model, and other more federated learning models on your phone. In this paper, we propose a client-task assignment strategy to preserve the fairness across multiple training tasks.

My research at HKUST: Medical AI

We proposed a pyramidally downsampled 3D Transformer to boost efficiency and maintain high segmentation accuracy. This paper also discussed the domain generalization problem, and proposed a cluster-based domain-adversarial learning method to exploit available domains at a fine-grained level.

Some Presentations

Teaching and Mentoring

Education

• M.S. in Electrical and Computer Engineering (Advanced Study), Carnegie Mellon University
  2023.08 - 2025.05, Pittsburgh, USA
  Current GPA: 4.0/4.0
• Exchange student in Electrical Engineering Department, The Technical University of Munich
  2023.04 - 2023.08, Munich, Germany
• B.Eng. in Automation (Advanced Class), Huazhong University of Science and Technology
  2019.09 - 2023.06, Wuhan, China
  GPA: 3.9/4.0

About Me

I was born in Xinxiang, China. My family name is 张 (Zhang), which is one of the most common family names in China. And my given name is 皓然 (Haoran), 皓(Hao) means "the color of the moon", 然(Ran) doesn't have a specific meaning, maybe it means "yeah, uh-huh". As the first in my family to attend college, I am deeply grateful for the support from my parents, mentors, and friends.

Template stolen from Jon Barron.
Last updated: