Approximate Representer Theorems in Non-reflexive Banach Spaces

Kevin Schlegel
Proceedings of the 31st International Conference on Algorithmic Learning Theory, PMLR 117:827-844, 2020.

Abstract

The representer theorem is one of the most important mathematical foundations for regularised learning and kernel methods. Classical formulations of the theorem state sufficient conditions under which a regularisation problem on a Hilbert space admits a solution in the subspace spanned by the representers of the data points. This turns the problem into an equivalent optimisation problem in a finite dimensional space, making it computationally tractable. Moreover, Banach space methods for learning have been receiving more and more attention. Considering the representer theorem in Banach spaces is hence of increasing importance. Recently the question of the necessary condition for a representer theorem to hold in Hilbert spaces and certain Banach spaces has been considered. It has been shown that a classical representer theorem cannot exist in general in non-reflexive Banach spaces. In this paper we propose a notion of approximate solutions and approximate representer theorem to overcome this problem. We show that for these notions we can indeed extend the previous results to obtain a unified theory for the existence of representer theorems in any general Banach spaces, in particular including $l^1$-type spaces. We give a precise characterisation when a regulariser admits a classical representer theorem and when only an approximate representer theorem is possible.

Cite this Paper


BibTeX
@InProceedings{pmlr-v117-schlegel20a, title = {Approximate Representer Theorems in Non-reflexive Banach Spaces}, author = {Schlegel, Kevin}, booktitle = {Proceedings of the 31st International Conference on Algorithmic Learning Theory}, pages = {827--844}, year = {2020}, editor = {Kontorovich, Aryeh and Neu, Gergely}, volume = {117}, series = {Proceedings of Machine Learning Research}, month = {08 Feb--11 Feb}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v117/schlegel20a/schlegel20a.pdf}, url = {https://proceedings.mlr.press/v117/schlegel20a.html}, abstract = {The representer theorem is one of the most important mathematical foundations for regularised learning and kernel methods. Classical formulations of the theorem state sufficient conditions under which a regularisation problem on a Hilbert space admits a solution in the subspace spanned by the representers of the data points. This turns the problem into an equivalent optimisation problem in a finite dimensional space, making it computationally tractable. Moreover, Banach space methods for learning have been receiving more and more attention. Considering the representer theorem in Banach spaces is hence of increasing importance. Recently the question of the necessary condition for a representer theorem to hold in Hilbert spaces and certain Banach spaces has been considered. It has been shown that a classical representer theorem cannot exist in general in non-reflexive Banach spaces. In this paper we propose a notion of approximate solutions and approximate representer theorem to overcome this problem. We show that for these notions we can indeed extend the previous results to obtain a unified theory for the existence of representer theorems in any general Banach spaces, in particular including $l^1$-type spaces. We give a precise characterisation when a regulariser admits a classical representer theorem and when only an approximate representer theorem is possible.} }
Endnote
%0 Conference Paper %T Approximate Representer Theorems in Non-reflexive Banach Spaces %A Kevin Schlegel %B Proceedings of the 31st International Conference on Algorithmic Learning Theory %C Proceedings of Machine Learning Research %D 2020 %E Aryeh Kontorovich %E Gergely Neu %F pmlr-v117-schlegel20a %I PMLR %P 827--844 %U https://proceedings.mlr.press/v117/schlegel20a.html %V 117 %X The representer theorem is one of the most important mathematical foundations for regularised learning and kernel methods. Classical formulations of the theorem state sufficient conditions under which a regularisation problem on a Hilbert space admits a solution in the subspace spanned by the representers of the data points. This turns the problem into an equivalent optimisation problem in a finite dimensional space, making it computationally tractable. Moreover, Banach space methods for learning have been receiving more and more attention. Considering the representer theorem in Banach spaces is hence of increasing importance. Recently the question of the necessary condition for a representer theorem to hold in Hilbert spaces and certain Banach spaces has been considered. It has been shown that a classical representer theorem cannot exist in general in non-reflexive Banach spaces. In this paper we propose a notion of approximate solutions and approximate representer theorem to overcome this problem. We show that for these notions we can indeed extend the previous results to obtain a unified theory for the existence of representer theorems in any general Banach spaces, in particular including $l^1$-type spaces. We give a precise characterisation when a regulariser admits a classical representer theorem and when only an approximate representer theorem is possible.
APA
Schlegel, K.. (2020). Approximate Representer Theorems in Non-reflexive Banach Spaces. Proceedings of the 31st International Conference on Algorithmic Learning Theory, in Proceedings of Machine Learning Research 117:827-844 Available from https://proceedings.mlr.press/v117/schlegel20a.html.

Related Material