Smoothness and uniqueness in ridge function representation

From MaRDI portal





A real-valued function defined on \(\mathbb{R}^n\) which can be written in the form \(f(a\cdot x)\), where \(a\) is a fixed non-zero vector and \(f:\mathbb{R}\to\mathbb{R}\), is called a ridge function. Let (1) \[ F(x)=\sum_{i=1}^m f_i (a^i\cdot x), \] where \(a^i\) are pairwise linearly independent. The author studies the following questions: If \(F\) is of a certain smoothness class, what can we say about the smoothness of the \(f_i\) ?...To what extent is a representation of \(F\) in the form (1) unique ? And, finally, which other ridge functions \(f(a\cdot x)\) can be written in the form (1) with \(a\neq\alpha a^i\), for any \(\alpha \in\mathbb{R}\) and \(i=1,...,m\) ? In the last section of the paper he discusses generalizations for sums of functions of the form \(f(Ax)\), where \(A\) is \(d\times n\) matrix and \(f:\mathbb{R}^d\to\mathbb{R}\).











This page was built for publication: Smoothness and uniqueness in ridge function representation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q740462)