support-vector-machinesHello, every one!
I don't know if someone can help me, but I should thank you for your attention.
I am now studying about generalization bound and I found that one of ways to estimate the bound is to essimate the leave one out error. This method is based on the theorem presented by A.Luntz and V.Brailovsky, which is in their paper: On the estimation of characters obtained in statistical procedure of recognition. It is pitty that this paper is written in RUSSIA and in 1969,33 years ago. I am hungering for reading the proof of that theorem. Does anyone happen to know this theorem's proof in english and will help me?
Thanks in advance!
wu tao
[log in to unmask]
2002-06-27
|