801 search results - page 5 / 161 » Machine Learning with Data Dependent Hypothesis Classes |
ECML
2003
Springer
14 years 20 days ago
2003
Springer
Abstract. Classical learning algorithms from the fields of artificial neural networks and machine learning, typically, do not take any costs into account or allow only costs depe...
ICMLA
2010
13 years 5 months ago
2010
Abstract--Learning multiple related tasks from data simultaneously can improve predictive performance relative to learning these tasks independently. In this paper we propose a nov...
ALT
2008
Springer
14 years 4 months ago
2008
Springer
It is investigated for which choice of a parameter q, denoting the number of contexts, the class of simple external contextual languages is iteratively learnable. On one hand, the ...
COLT
2004
Springer
14 years 25 days ago
2004
Springer
ICML
2006
IEEE
14 years 8 months ago
2006
IEEE
We formalize the associative bandit problem framework introduced by Kaelbling as a learning-theory problem. The learning environment is modeled as a k-armed bandit where arm payof...