Sciweavers

ECML
2007
Springer
14 years 3 months ago
Structure Learning of Probabilistic Relational Models from Incomplete Relational Data
Abstract. Existing relational learning approaches usually work on complete relational data, but real-world data are often incomplete. This paper proposes the MGDA approach to learn...
Xiao-Lin Li, Zhi-Hua Zhou
ECML
2007
Springer
14 years 3 months ago
Class Noise Mitigation Through Instance Weighting
We describe a novel framework for class noise mitigation that assigns a vector of class membership probabilities to each training instance, and uses the confidence on the current ...
Umaa Rebbapragada, Carla E. Brodley
ECML
2007
Springer
14 years 3 months ago
Dual Strategy Active Learning
Abstract. Active Learning methods rely on static strategies for sampling unlabeled point(s). These strategies range from uncertainty sampling and density estimation to multi-factor...
Pinar Donmez, Jaime G. Carbonell, Paul N. Bennett
ECML
2007
Springer
14 years 3 months ago
Statistical Debugging Using Latent Topic Models
Abstract. Statistical debugging uses machine learning to model program failures and help identify root causes of bugs. We approach this task using a novel Delta-Latent-Dirichlet-Al...
David Andrzejewski, Anne Mulhern, Ben Liblit, Xiao...
ECML
2007
Springer
14 years 3 months ago
Finding the Right Family: Parent and Child Selection for Averaged One-Dependence Estimators
Averaged One-Dependence Estimators (AODE) classifies by uniformly aggregating all qualified one-dependence estimators (ODEs). Its capacity to significantly improve naive Bayes...
Fei Zheng, Geoffrey I. Webb
ECML
2007
Springer
14 years 3 months ago
A Simple Lexicographic Ranker and Probability Estimator
Given a binary classification task, a ranker sorts a set of instances from highest to lowest expectation that the instance is positive. We propose a lexicographic ranker, LexRank,...
Peter A. Flach, Edson Takashi Matsubara
ECML
2007
Springer
14 years 3 months ago
Fast Optimization Methods for L1 Regularization: A Comparative Study and Two New Approaches
L1 regularization is effective for feature selection, but the resulting optimization is challenging due to the non-differentiability of the 1-norm. In this paper we compare state...
Mark Schmidt, Glenn Fung, Rómer Rosales