Rate My Professor Karl Rohe

Karl Rohe is a Professor of Statistics at the University of Wisconsin-Madison, with courtesy appointments in the School of Journalism and Mass Communication, the Department of Electrical and Computer Engineering, and the Department of Educational Psychology. He earned his Ph.D. in Statistics from the University of California, Berkeley, in May 2011, with a designated emphasis in Communication, Computation, and Statistics. His doctoral thesis, titled "Analysis of Spectral Clustering and the Lasso under Nonstandard Statistical Models," was supervised by prominent statisticians. Prior to his doctorate, Rohe received a B.S. in Statistics, summa cum laude, from Michigan State University in 2006, along with a minor in Environmental Economics. He began his academic career at the University of Wisconsin-Madison as an Assistant Professor in the Department of Statistics in September 2011, advanced to Associate Professor with tenure in September 2017, and was subsequently promoted to Full Professor.

Rohe's research centers on creating statistical tools and theoretical frameworks for analyzing large social graphs, emphasizing network sampling methods, spectral clustering algorithms, and matrix factorization techniques tailored to the data science era. His seminal contributions include the paper "Spectral Clustering and the High-Dimensional Stochastic Blockmodel" (with S. Chatterjee and B. Yu, 2011), which has garnered over 1,200 citations and advanced clustering in high-dimensional networks; "Regularized Spectral Clustering under the Degree-Corrected Stochastic Blockmodel" (with T. Qin, Advances in Neural Information Processing Systems, 2013); "A Critical Threshold for Design Effects in Network Sampling" (The Annals of Statistics, 2019); and "Asymptotic Seed Bias in Respondent-Driven Sampling" (Electronic Journal of Statistics, 2020). Rohe received the Evelyn Fix Memorial Medal and Citation in 2011 for demonstrating the greatest promise in statistical research at UC Berkeley. He serves as an Associate Editor for the Journal of the Royal Statistical Society: Series B and previously for the Journal of the American Statistical Association. Rohe has delivered invited talks at prestigious venues such as the Joint Statistical Meetings, NeurIPS, and international workshops, including presentations on spectral techniques for directed networks and transitivity in stochastic blockmodels. His work bridges classical methods like principal component analysis with modern embeddings in machine learning, significantly impacting network analysis, respondent-driven sampling, and statistical inference for massive datasets.