$\ell_p$ Slack Norm Support Vector Data Description

TL;DR

This paper extends the SVDD method for one-class classification by introducing an $ ext{ell}_p$-norm slack penalty, allowing for non-linear cost functions and improved sparsity, with theoretical and experimental validation.

Contribution

It generalizes SVDD to $ ext{ell}_p$-norm slack penalties, enabling non-linear costs and better sparsity control, with theoretical analysis and empirical validation.

Findings

Enhanced descriptive capacity due to sparsity-inducing dual norm

Theoretical bounds on generalization performance based on $p$

Experimental results show improved performance over existing methods

Abstract

The support vector data description (SVDD) approach serves as a de facto standard for one-class classification where the learning task entails inferring the smallest hyper-sphere to enclose target objects while linearly penalising any errors/slacks via an $\ell_1$-norm penalty term. In this study, we generalise this modelling formalism to a general $\ell_p$-norm ($p\geq1$) slack penalty function. By virtue of an $\ell_p$ slack norm, the proposed approach enables formulating a non-linear cost function with respect to slacks. From a dual problem perspective, the proposed method introduces a sparsity-inducing dual norm into the objective function, and thus, possesses a higher capacity to tune into the inherent sparsity of the problem for enhanced descriptive capability. A theoretical analysis based on Rademacher complexities characterises the generalisation performance of the proposed approach in terms of parameter $p$ while the experimental results on several datasets confirm the merits of the proposed method compared to other alternatives.