The overarching goal of his research is to improve predictive data science and scientific computing via intelligent computation.

Haizhao YangFrom navigating with map apps and streaming new music to virtual personal assistants like Alexa and Siri, machine learning is a valuable tool that has become an intrinsic part of our daily lives. Yet these systems are not always reliable because they sometimes provide inaccurate information.

Haizhao Yang, an associate professor in the University of Maryland’s Department of Mathematics, is paving the way for more dependable machine learning systems by advancing its subset of deep learning. Deep learning is a method that uses three or more layers of neural networks to learn information and has led to tremendous breakthroughs in the field by providing more interpretable and precise outcomes.

The overarching goal of his research is to improve predictive data science and scientific computing via intelligent computation, which could have significant impacts on several fields, like weather prediction and health care.

Because deep learning is still in its infancy, establishing mathematical and statistical principles is an important step in improving its ability to obtain reliable results in these applications, Yang explained.

However, there are several challenges that he must overcome to make deep learning more reliable. The first obstacle is gaining a better understanding of how physical processes, like weather phenomena for example, can be applied to his research.


The second challenge involves lowering the expensive computational costs associated with the creation of deep learning. In order to overcome these challenges, Yang will utilize artificial intelligence to examine historical data and develop numerical strategies to make fast computations.


Yang, who recently started a new affiliate appointment at the University of Maryland Institute for Advanced Computer Studies (UMIACS), is looking forward to collaborating with fellow faculty members—especially those in the University of Maryland Center for Machine Learning—and utilizing the institute’s state-of-the-art computing infrastructure.

“The unique, interactive environment and powerful resources provided by UMIACS will accelerate my research on advancing intelligent computation,” he explained.

Yang has received several prestigious awards, including the award for Maryland Research Excellence in 2023, the Office of Naval Research Young Investigator Award in 2022, the Teaching for Tomorrow Award at Purdue University in 2021 and a National Science Foundation CAREER Award in 2020.

Before coming to UMD, Yang was an assistant professor of mathematics at Purdue University. He also held an affiliate position at the Institute of Data Science at the National University of Singapore. He earned his Ph.D. in mathematics from Stanford University in 2015.


Adapted from an article written by Ethan Cannistra, UMIACS communications group

  • 1
  • 2
  • 3
  • 4