Abstract:The necessary and sufficient conditions for existence of a generalized representer theorem are presented for learning Hilbert space-valued functions. Representer theorems involving explicit basis functions and Reproducing Kernels are a common occurrence in various machine learning algorithms like generalized least squares, support vector machines, Gaussian process regression and kernel based deep neural networks to name a few. Due to the more general structure of the underlying variational problems, the theory is also relevant to other application areas like optimal control, signal processing and decision making. We present the generalized representer as a unified view for supervised and semi-supervised learning methods, using the theory of linear operators and subspace valued maps. The implications of the theorem are presented with examples of multi input-multi output regression, kernel based deep neural networks, stochastic regression and sparsity learning problems as being special cases in this unified view.

The Representation and Computational Procedures for the Generalized Inverse [image omitted] of an Operator A in Hilbert Spaces

The Representation and Computational Procedures for the Generalized Inverse of an Operator<i>A</i>in Hilbert Spaces

The Representation and Approximations of Outer Generalized Inverses

The Representation and Approximation of the Drazin Inverse of a Linear Operator in Hilbert Space

The Representation And Approximation For The Generalized Inverse A(T,S)((2))

Representations for the Drazin Inverse of Bounded Operators on Banach Space

The Characterizations and Representations for the Generalized Inverses with Prescribed Idempotents in Banach Algebras

The Quasi-Linear Operator Outer Generalized Inverse with Prescribed Range and Kernel in Banach Spaces

Representations For The Generalized Drazin Inverse Of Bounded Linear Operators

Integral and limit representations of the outer inverse in Banach space

Operators with Equal Projections Related to Their Generalized Inverses.

Perturbation Analysis of Bounded Homogeneous Generalized Inverses on Banach Spaces

The Representation and Approximation of the W-weighted Drazin Inverse of Linear Operators in Hilbert Space

Representations for Moore-Penrose inverses in Hilbert spaces

A characterization and representation of the generalized inverse A(2)T,S and its applications

The expression of the generalized inverse of the perturbed operator under Type I perturbation in Hilbert spaces

On a New Generalized Inverse for Hilbert Space Operators

A Generalized Representer Theorem for Hilbert Space - Valued Functions

On Integral Representation Of The Generalized Inverse A(T),((2))(S)

Determinantal representation of the generalized inverse \bi A_{T,S}^{(2)} over integral domains and its applications

The Sherman-Morrison-Woodbury Formula for the Generalized Inverses