A Nonconformity Approach to Model Selection for SVMs

TL;DR

This paper introduces a nonconformity-based training algorithm for SVMs that eliminates the need for traditional model selection methods like cross-validation, providing theoretical guarantees and faster convergence.

Contribution

It presents a novel nonconformity measure-based algorithm for SVM model selection, avoiding cross-validation and offering theoretical bounds on generalization error.

Findings

Comparable accuracy to standard methods

Theoretical generalization error bounds

Faster convergence in training

Abstract

We investigate the issue of model selection and the use of the nonconformity (strangeness) measure in batch learning. Using the nonconformity measure we propose a new training algorithm that helps avoid the need for Cross-Validation or Leave-One-Out model selection strategies. We provide a new generalisation error bound using the notion of nonconformity to upper bound the loss of each test example and show that our proposed approach is comparable to standard model selection methods, but with theoretical guarantees of success and faster convergence. We demonstrate our novel model selection technique using the Support Vector Machine.