Sciweavers

225 search results - page 20 / 45
» On Learning Decision Trees with Large Output Domains
Sort
View
ICML
2000
IEEE
14 years 8 months ago
Using Error-Correcting Codes for Text Classification
This paper explores in detail the use of Error Correcting Output Coding (ECOC) for learning text classifiers. We show that the accuracy of a Naive Bayes Classifier over text class...
Rayid Ghani
JSA
2006
97views more  JSA 2006»
13 years 7 months ago
Dynamic feature selection for hardware prediction
It is often possible to greatly improve the performance of a hardware system via the use of predictive (speculative) techniques. For example, the performance of out-of-order micro...
Alan Fern, Robert Givan, Babak Falsafi, T. N. Vija...
KDD
2006
ACM
155views Data Mining» more  KDD 2006»
14 years 8 months ago
Camouflaged fraud detection in domains with complex relationships
We describe a data mining system to detect frauds that are camouflaged to look like normal activities in domains with high number of known relationships. Examples include accounti...
Sankar Virdhagriswaran, Gordon Dakin
ICMLA
2010
13 years 5 months ago
Boosting Multi-Task Weak Learners with Applications to Textual and Social Data
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...
Jean Baptiste Faddoul, Boris Chidlovskii, Fabien T...
ECML
2006
Springer
13 years 11 months ago
Bandit Based Monte-Carlo Planning
Abstract. For large state-space Markovian Decision Problems MonteCarlo planning is one of the few viable approaches to find near-optimal solutions. In this paper we introduce a new...
Levente Kocsis, Csaba Szepesvári