Timezone: »

 
Poster
Asymmetric Valleys: Beyond Sharp and Flat Local Minima
Haowei He · Gao Huang · Yang Yuan

Tue Dec 10 05:30 PM -- 07:30 PM (PST) @ East Exhibition Hall B + C #116
in Optimization -- Non-Convex Optimization »

Despite the non-convex nature of their loss functions, deep neural networks are known to generalize well when optimized with stochastic gradient descent (SGD). Recent work conjectures that SGD with proper configuration is able to find wide and flat local minima, which are correlated with good generalization performance. In this paper, we observe that local minima of modern deep networks are more than being flat or sharp. Instead, at a local minimum there exist many asymmetric directions such that the loss increases abruptly along one side, and slowly along the opposite side – we formally define such minima as asymmetric valleys. Under mild assumptions, we first prove that for asymmetric valleys, a solution biased towards the flat side generalizes better than the exact empirical minimizer. Then, we show that performing weight averaging along the SGD trajectory implicitly induces such biased solutions. This provides theoretical explanations for a series of intriguing phenomena observed in recent work [25, 5, 51]. Finally, extensive empirical experiments on both modern deep networks and simple 2 layer networks are conducted to validate our assumptions and analyze the intriguing properties of asymmetric valleys.

Keywords: Non-Convex Optimization · Optimization for Deep Networks · Deep Learning · Optimization

Author Information

Haowei He (Tsinghua University)
Gao Huang (Tsinghua)
Yang Yuan (Cornell University)

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

More from the Same Authors