Search papers, labs, and topics across Lattice.
Peking University
1
0
2
Momentum *can* accelerate single-pass stochastic gradient descent for generalized linear prediction, resolving a long-standing open question and outperforming variance reduction techniques.
