2988 search results - page 30 / 598 » Experiments with a New Boosting Algorithm |
ICML
2006
IEEE
16 years 18 days ago
2006
IEEE
A multiclass classification problem can be reduced to a collection of binary problems with the aid of a coding matrix. The quality of the final solution, which is an ensemble of b...
AAAI
1998
15 years 1 months ago
1998
Unpredictability in the running time of complete search procedures can often be explained by the phenomenon of "heavy-tailed cost distributions", meaning that at any tim...
AAAI
2006
15 years 1 months ago
2006
Many learning tasks in adversarial domains tend to be highly dependent on the opponent. Predefined strategies optimized for play against a specific opponent are not likely to succ...
CIKM
2008
Springer
15 years 1 months ago
2008
Springer
This paper points out that many machine learning problems in IR should be and can be formalized in a novel way, referred to as `group-based learning'. In group-based learning...
CIKM
2005
Springer
15 years 1 months ago
2005
Springer
This paper presents a new pooling method for constructing the assessment sets used in the evaluation of retrieval systems. Our proposal is based on RankBoost, a machine learning v...