The generalization error of max-margin linear classifiers: Benign overfitting and high dimensional asymptotics in the overparametrized regime
Andrea Montanari Thanks: Department of Electrical Engineering and Department of Statistics, Stanford University Feng Ruan Thanks: Department of Statistics and Data Science, Northwestern University Youngtak Sohn Thanks: Department of Mathematics, Massachusetts Institute of Technology Jun Yan Thanks: Department of Statistics, Stanford University
Abstract
Modern machine learning classifiers often exhibit vanishing classification error on the training set. They achieve this by learning nonlinear representations of the inputs that maps the data into linearly separable classes.
原文 arXiv:1911.01544;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1911.01544v3