Notes on Hierarchical Splines, DCLNs and i-theory

TitleNotes on Hierarchical Splines, DCLNs and i-theory
Publication TypeCBMM Memo
Year of Publication2015
AuthorsPoggio T., L. R, Shashua A., Cohen N., Anselmi F.
NumberCBMM Memo No. 037

We define an extension of classical additive splines for multivariate
function approximation that we call hierarchical splines. We show that the
case of hierarchical, additive, piece-wise linear splines includes present-day
Deep Convolutional Learning Networks (DCLNs) with linear rectifiers and
pooling (sum or max). We discuss how these observations together with
i-theory may provide a framework for a general theory of deep networks.

Citation Key7

CBMM Memo No.: