ICML
2007
IEEE
15 years 9 days ago
2007
IEEE
A general framework is proposed for gradient boosting in supervised learning problems where the loss function is defined using a kernel over the output space. It extends boosting ...