Sciweavers

81
Voted
IJCNN
2007
IEEE
15 years 4 months ago
Optimizing 0/1 Loss for Perceptrons by Random Coordinate Descent
—The 0/1 loss is an important cost function for perceptrons. Nevertheless it cannot be easily minimized by most existing perceptron learning algorithms. In this paper, we propose...
Ling Li, Hsuan-Tien Lin
79
Voted
ICML
2009
IEEE
15 years 10 months ago
Boosting with structural sparsity
Despite popular belief, boosting algorithms and related coordinate descent methods are prone to overfitting. We derive modifications to AdaBoost and related gradient-based coordin...
John Duchi, Yoram Singer