Abstract: Algorithmic stability for regression and classification
In a supervised learning setting, a model fitting algorithm is unstable if small perturbations to the input (the training data) can often lead to large perturbations in the output (say, predictions returned by the fitted model). Algorithmic stability is a desirable property with many important implications such as generalization and robustness, but testing the stability property empirically is known to be impossible in the setting of complex black-box models. In this work, we establish that bagging any black-box regression algorithm automatically ensures that stability holds, with no assumptions on the algorithm or the data. Furthermore, we construct a new framework for defining stability in the context of classification, and show that using bagging to estimate our uncertainty about the output label will again allow stability guarantees for any black-box model. This work is joint with Jake Soloff and Rebecca Willett. Bio: Professor in the Department of Statistics at the University of Chicago. Before starting at U of C, I was a NSF postdoctoral fellow during 2012-13 in the Department of Statistics at Stanford University, supervised by Emmanuel Candès. I received my PhD in Statistics at the University of Chicago in 2012, advised by Mathias Drton and Nati Srebro, and a MS in Mathematics at the University of Chicago in 2009. Prior to graduate school, I was a mathematics teacher at the Park School of Baltimore from 2005 to 2007, and received an ScB in Mathematics from Brown University in 2005.