Covering Number Bounds of Certain Regularized Linear Function Classes

Recently, sample complexity bounds have been derived for problems involving linear functions such as neural networks and support vector machines. In many of these theoretical studies, the concept of covering numbers played an important role. It is thus useful to study covering numbers for linear function classes. In this paper, we investigate two closely related methods to derive upper bounds on these covering numbers. The first method, already employed in some earlier studies, relies on the so-called Maurey's lemma; the second method uses techniques from the mistake bound framework in online learning. We compare results from these two methods, as well as their consequences in some learning formulations.

• Type: Journal
• Year: 2002
• Where: JMLR
• Authors: Tong Zhang
• Keywords: JMLR 2002, Linear Function, Linear Function Classes, Sample Complexity