Poster

Provably Strict Generalisation Benefit for Equivariant Models

Bryn Elesedy · Sheheryar Zaidi

Virtual

Keywords: [ Architectures ] [ Visual Scene Analysis and Interpretation ] [ Applications -> Object Detection; Neuroscience and Cognitive Science ] [ Visual Perception ]

[ Abstract ]
[ Slides
[ Paper ]
[ Visit Poster at Spot C6 in Virtual World ]
Tue 20 Jul 9 a.m. PDT — 11 a.m. PDT
 
Spotlight presentation: AutoML and Deep Architecture
Tue 20 Jul 6 a.m. PDT — 7 a.m. PDT

Abstract:

It is widely believed that engineering a model to be invariant/equivariant improves generalisation. Despite the growing popularity of this approach, a precise characterisation of the generalisation benefit is lacking. By considering the simplest case of linear models, this paper provides the first provably non-zero improvement in generalisation for invariant/equivariant models when the target distribution is invariant/equivariant with respect to a compact group. Moreover, our work reveals an interesting relationship between generalisation, the number of training examples and properties of the group action. Our results rest on an observation of the structure of function spaces under averaging operators which, along with its consequences for feature averaging, may be of independent interest.

Chat is not available.