Strong Convex Loss Can Increase the Learning Rates of Online Learning.

Baohuai Sheng Liqin Duan Pei-Xin Ye

Published in: J. Comput. (2014)

Keyphrases

online learning
learning rate
learning algorithm
convergence rate
covering numbers
gaussian kernels
active learning
convergence speed
e learning
uniform convergence
decision trees
high dimensional
sufficient conditions
convex optimization
regret bounds