Poster
How SGD Selects the Global Minima in Over-parameterized Learning: A Dynamical Stability Perspective
Lei Wu · Chao Ma · Weinan E
Room 210 #29
Keywords: [ Learning Theory ] [ Optimization ] [ Dynamical Systems ] [ Optimization for Deep Networks ]
The question of which global minima are accessible by a stochastic gradient decent (SGD) algorithm with specific learning rate and batch size is studied from the perspective of dynamical stability. The concept of non-uniformity is introduced, which, together with sharpness, characterizes the stability property of a global minimum and hence the accessibility of a particular SGD algorithm to that global minimum. In particular, this analysis shows that learning rate and batch size play different roles in minima selection. Extensive empirical results seem to correlate well with the theoretical findings and provide further support to these claims.
Live content is unavailable. Log in and register to view live content