Surrogate Losses for Online Learning of Stepsizes in Stochastic Non-Convex Optimization
zhenxun zhuang · Ashok Cutkosky · Francesco Orabona

Wed Jun 12th 02:35 -- 02:40 PM @ Room 104

Stochastic Gradient Descent (SGD) has played a central role in machine learning. However, it requires a carefully hand-picked stepsize for fast convergence, which is notoriously tedious and time-consuming to tune. Over the last several years, a plethora of adaptive gradient-based algorithms have emerged to ameliorate this problem. In this paper, we propose new surrogate losses to cast the problem of learning the optimal stepsizes for the stochastic optimization of a non-convex smooth objective function onto an online convex optimization problem. This allows the use of no-regret online algorithms to compute optimal stepsizes on the fly. In turn, this results in a SGD algorithm with self-tuned stepsizes that guarantees convergence rates that are automatically adaptive to the level of noise.

Author Information

zhenxun zhuang (Boston University)
Ashok Cutkosky (Google)
Francesco Orabona (Stony Brook University)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors