Sharper lower bounds on the performance of the empirical risk minimization algorithm

DOI10.3150/09-BEJ225zbMath1220.62007arXiv1102.4983OpenAlexW3098804594MaRDI QIDQ637070

Shahar Mendelson, Guillaume Lecué

Publication date: 2 September 2011

Published in: Bernoulli (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1102.4983

zbMATH Keywords

learning theorylower bounduniform central limit theoremmultidimensional central limit theorem

Mathematics Subject Classification ID

Gaussian processes (60G15) Central limit and other weak theorems (60F05) Inference from stochastic processes (62M99) Statistical decision theory (62C99)

Related Items

Performance of empirical risk minimization in linear aggregation, Empirical risk minimization is optimal for the convex aggregation problem, General oracle inequalities for model selection, On the optimality of the aggregate with exponential weights for low temperatures, General nonexact oracle inequalities for classes with a subexponential envelope, Optimal learning with \textit{Q}-aggregation

Cites Work

• Risk bounds for statistical learning
• Weak convergence and empirical processes. With applications to statistics
• Local Rademacher complexities and oracle inequalities in risk minimization. (2004 IMS Medallion Lecture). (With discussions and rejoinder)
• Empirical minimization
• Lower Bounds for the Empirical Minimization Algorithm
• Uniform Central Limit Theorems
• The Generic Chaining
• The importance of convexity in learning with squared loss
• Suboptimality of Penalized Empirical Risk Minimization in Classification