L4: Practical loss-based stepsize adaptation for deep learning.

Michal Rolinek Georg Martius

Published in: CoRR (2018)

Keyphrases

deep learning
step size
unsupervised learning
convergence rate
unsupervised feature learning
machine learning
approximate dynamic programming
cost function
mental models
weakly supervised
convergence speed
neural network
multiscale
reinforcement learning
decision makers