Separating Geometry from Probability in the Analysis of Generalization

Maxim Raginsky, Benjamin Recht

The goal of machine learning is to find models that minimize prediction error on data that has not yet been seen. Its operational paradigm assumes access to a dataset $S$ and artic…