Analysis of Perceptron-Based Active Learning

We start by showing that in an active learning setting, the Perceptron algorithm needs $\Omega(\frac{1}{\epsilon^2})$ labels to learn linear separators within generalization error $\epsilon$. We then present a simple selective sampling algorithm for this problem, which combines a modification of th...

Descripció completa

Dades bibliogràfiques
Autors principals: Dasgupta, Sanjoy, Kalai, Adam Tauman, Monteleoni, Claire
Idioma:en_US
Publicat: 2005
Matèries:
Accés en línia:http://hdl.handle.net/1721.1/30585