Dr Archer Gong Zhang he/him
Developing statistical theory and methodology for learning and inference across heterogeneous data and distributions
Lecturer in Statistics and Data Analytics
School of Mathematics and Statistics
Research interests:
Statistical Methodology, Statistical Theory, Semiparametric Inference, Nonparametric Inference, Causal Inference, Data Integration and Fusion, Distributional Inference, Distribution Shift, Robust Statistics, Statistical Learning
Research fields:
Statistics, Data Science, Artificial Intelligence and Machine Learning, Environmental and Climate Science, Health and Biomedical Science
Strategic Research Areas
Why do you want to join the DiveIn community?
Much of my research is motivated by methodological questions that arise when combining complex and heterogeneous data sources. I would like to join DiveIn to develop these ideas in genuinely interdisciplinary settings, where statistical methodology can be shaped by substantive problems and, in turn, contribute to areas such as health, environmental science and AI. I am also interested in co-supervising PhD students across disciplinary boundaries.
Personal profile:
I am a statistician interested in developing new methodology for complex, heterogeneous and imperfect data, supported by rigorous statistical theory. Much of my research is motivated by questions that arise when information comes from heterogeneous but related populations, studies or data-generating mechanisms. My work spans semiparametric and nonparametric inference, data integration, causal inference, statistical learning under distribution shift, and robust statistics, with a particular interest in methods that make efficient use of information shared across data sources while remaining flexible about the underlying distributions.
I am particularly interested in collaborations where scientific problems lead to new statistical questions. My current and emerging collaborations involve health and biomedical applications, environmental uncertainty, statistical learning and dynamical systems. Through DiveIn, I would like to work with researchers and students from other disciplines on problems where statistical methodology is an important part of the research.
Possible PhD projects could involve combining information from heterogeneous data sources, learning under distribution shift, uncertainty quantification in environmental or health applications, or developing distributional and causal methods for complex data. I would be especially interested in supervising students who enjoy methodological work but are also willing to engage seriously with an application area. My supervision style is to give students room to develop their own ideas while maintaining regular discussions and providing clear technical guidance when needed. I also think it is just as important to discuss approaches that do not work, since understanding why they fail often leads to better questions.
In teaching and supervision, I have worked with students from a wide range of academic and cultural backgrounds. I try to set clear expectations, make it easy for students to ask questions, and adjust the level and style of support to their experience and goals. I see this as an important part of creating a research environment in which students can develop confidence and independence.
One of the things I enjoy most about research is learning enough about a new scientific area to identify the statistical problem underneath it. I like projects where this leads to both a useful methodological contribution and a better solution to the original applied problem.
