The Bounds on the Rate of Uniform Convergence of Learning Process with Rough Samples

摘要

Support vector machine is a research hotspot in the area of machine learning, and the bounds on the rate of uniform convergence of statistical learning theory describe the extended ability of learning machine based on ERM. In the paper, Rough Empirical Risk Minimization (RERM) principle is proposed, and the bounds on the rate of uniform convergence of learning process with rough samples are presented and proven, they provide a theoretical basis for the research of rough support vector machine. Which has a wide range of applications in Natural Language Processing, including automatic summarization, text classification, etc.