Alden Green, an assistant professor in the Department of Applied Mathematics and Statistics, came to JHU from Stanford University, where he was a postdoctoral Stein Fellow. In this Q&A, he shares his interest in statistics, his work uncovering structure in complex data, the real-world applications he’s exploring, and why he’s excited to pursue this research at Johns Hopkins.
Tell us a little about yourself.
I grew up in Montgomery County, Maryland—a suburb of D.C.—so starting at JHU feels like a homecoming for me. As a kid, I was a big Orioles fan and obsessed with the statistics of sports; I use to make my mom copy Orioles box scores at breakfast, and later I convinced my family buy some regression software (I have no memory of what it was) so I could build models predicting NCAA tournament results. I lost interest in this as an undergraduate at Harvard where I’d planned to study history or economics, until I took Stats 110, an introductory probability course. By the time we finished combinatorics, I knew statistics was where I belonged.
After graduating, I started a PhD program in the statistics department at Carnegie Mellon. Six wonderful years followed—featuring a little sports statistics and, thankfully, no combinatorics—during which time I developed a passion for research and collaboration in the statistics field.
Describe your research.
My research spans several areas, but the common thread is developing theory and methods for making sense of complex, high-dimensional data—essentially, situations where we have far more variables than we can easily analyze. I’m especially interested in uncovering and using low-dimensional patterns hidden within that complexity.
Many modern datasets are enormous—for example, in genomics, we might measure hundreds of thousands of variables for each individual. Statistical theory tells us that analyzing data on this scale can be extremely difficult, or even impossible, unless there’s some simpler underlying structure. My work uses statistical theory to understand how algorithms can detect and take advantage of that hidden structure. In one line of research, for instance, I’ve shown that certain nonlinear dimensionality reduction methods—those based on graph Laplacians—can optimally capture low-dimensional “manifold” structures in data, leading to much better performance than would otherwise be possible.
What are some of the real-world applications of your research?
One area I’m working in that has significant applied potential is selective inference for peaks of random fields. Random fields often model data in spatiotemporal settings such as neuroimaging, where peaks can correspond to regions of the brain that activate in response to a stimulus. Conducting inference for these peaks can help neuroscientists answer questions like: “Is the activity we’re seeing in this region of the brain statistically significant?”, “Where is the response the largest?”, and “How large is the effect size?”
What drew you to this field and focus area?
I’m drawn to problems that are motivated by real applications but still allow for clear rigorous theoretical analysis. In particular, I enjoy working on questions that have well-defined, theoretically optimal solutions.
What excites you about bringing this work to Johns Hopkins?
Johns Hopkins is making a remarkable investment in data science, statistics, machine learning, and AI, making this an especially exciting time to be doing statistics research at the university. I’m eager to help strengthen an interdisciplinary culture among researchers working in these areas. More broadly, Johns Hopkins is an incredible applied scientific community, and I’m looking forward to learning about cutting-edge problems in fields like biomedicine and exploring how statisticians can contribute to solving them.
What are some of your goals for this first year at JHU?
First and foremost, I’m hoping to get to know everybody in AMS, and I encourage colleagues and students I haven’t met yet to feel welcome to stop by my office. Beyond that, I’d like to start some applied collaborations within the department and across the broader university community.
Anything else we should know? Any fun facts?
I’m very outdoorsy—I enjoy hiking, running, biking, playing soccer, and taking long, unhurried walks in the afternoons. A perhaps less conventional fact about me is that I like to read challenging books, such as Infinite Jest or Gravity’s Rainbow, aloud to myself—occasionally even in coffee shops, where I do my best to stay focused and avoid drawing curious glances.