Timezone: »

Convex Deep Learning via Normalized Kernels
Özlem Aslan · Xinhua Zhang · Dale Schuurmans

Wed Dec 10 04:00 PM -- 08:59 PM (PST) @ Level 2, room 210D

Deep learning has been a long standing pursuit in machine learning, which until recently was hampered by unreliable training methods before the discovery of improved heuristics for embedded layer training. A complementary research strategy is to develop alternative modeling architectures that admit efficient training methods while expanding the range of representable structures toward deep models. In this paper, we develop a new architecture for nested nonlinearities that allows arbitrarily deep compositions to be trained to global optimality. The approach admits both parametric and nonparametric forms through the use of normalized kernels to represent each latent layer. The outcome is a fully convex formulation that is able to capture compositions of trainable nonlinear layers to arbitrary depth.

Author Information

Özlem Aslan (University of Alberta)
Xinhua Zhang (University of Illinois at Chicago (UIC))
Dale Schuurmans (Google Brain & University of Alberta)

More from the Same Authors