801 search results - page 37 / 161 » The Inefficiency of Batch Training for Large Training Sets |
LREC
2008
13 years 10 months ago
2008
This paper describes the creation of a state-of-the-art answer type detection system capable of recognizing more than 200 different expected answer types with greater than 85% pre...
LWA
2007
13 years 10 months ago
2007
The goal of this paper is to investigate to what extent a rule learning heuristic can be learned from experience. Our basic approach is to learn a large number of rules and record ...
CRV
2009
IEEE
14 years 3 months ago
2009
IEEE
In this paper we introduce the concept and method for adaptively tuning the model complexity in an online manner as more examples become available. Challenging classification pro...
NIPS
2003
13 years 10 months ago
2003
Label ranking is the task of inferring a total order over a predefined set of labels for each given instance. We present a general framework for batch learning of label ranking f...
ICML
2003
IEEE
14 years 9 months ago
2003
IEEE
We present a fast iterative support vector training algorithm for a large variety of different formulations. It works by incrementally changing a candidate support vector set usin...