A Generalized Representer Theorem for Hilbert Space - Valued Functions

TL;DR

This paper establishes necessary and sufficient conditions for a generalized representer theorem applicable to Hilbert space-valued functions, unifying various machine learning and signal processing methods under a common theoretical framework.

Contribution

It introduces a generalized representer theorem for Hilbert space-valued functions, encompassing supervised, semi-supervised learning, and other applications using linear operators and subspace maps.

Findings

Unified view of supervised and semi-supervised learning methods

Applicable to multi-input multi-output regression and neural networks

Provides theoretical foundation for sparsity and stochastic regression

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.