Can I read One-Class SVM Bayes Consistency Analysis on EtoBox?

One-Class SVM Bayes Consistency Analysis by TungKVT is a document available to read on EtoBox.

What is One-Class SVM Bayes Consistency Analysis about?

This document summarizes the main results of a paper about the consistency of one-class SVM and related algorithms. It shows that these algorithms can achieve Bayes consistency, even when the regularization parameter is held fixed, as long as the bandwidth of the Gaussian kernel decreases appropriately as the number of samples increases. Specifically, it proves that the classification error converges to the Bayes risk and the decision function converges to a truncated density estimate. This establishes the

Author

TungKVT

Language

EN