Timezone: »

Path-SGD: Path-Normalized Optimization in Deep Neural Networks
Behnam Neyshabur · Russ Salakhutdinov · Nati Srebro

Tue Dec 08 04:00 PM -- 08:59 PM (PST) @ 210 C #33 #None

We revisit the choice of SGD for training deep neural networks by reconsidering the appropriate geometry in which to optimize the weights. We argue for a geometry invariant to rescaling of weights that does not affect the output of the network, and suggest Path-SGD, which is an approximate steepest descent method with respect to a path-wise regularizer related to max-norm regularization. Path-SGD is easy and efficient to implement and leads to empirical gains over SGD and AdaGrad.

Author Information

Behnam Neyshabur (TTI Chicago)
Russ Salakhutdinov (University of Toronto)
Nati Srebro (Toyota Technological Institute at Chicago)

More from the Same Authors