Sequential Minimal Optimization Algorithm for One-Class Support Vector Machines With Privileged Information

2026-06-20

AI Summary

This work addresses the challenge of modeling one-class support vector machines (OC-SVMs) under the learning using privileged information (LUPI) paradigm, where privileged information is available during training but inaccessible at test time. The paper presents the first sequential minimal optimization (SMO) algorithm tailored for OC-SVM within the LUPI framework. By employing an efficient two-variable optimization strategy, the proposed method enables scalable training while guaranteeing convergence in a finite number of steps, as rigorously proven. Furthermore, it elucidates how privileged information shapes the anomaly boundary in the original feature space. Experimental results on multiple benchmark datasets demonstrate that the algorithm significantly outperforms non-sequential approaches such as interior-point methods, achieving superior performance in both training efficiency and model accuracy.

Abstract

One of the powerful techniques in data modeling is accounting for features that are available at the training stage, but are not available when the trained model is used to classify or predict test data -- the Learning Using Privileged Information paradigm (LUPI). Sequential Minimal Optimization (SMO) methods have been developed for supervised Support Vector Machines (SVM), unsupervised one-class SVM, and SVM with privileged information (SVM+). The missing brick in this research has long been a one-class SVM with privileged information (OC-SVM+). In this paper, we propose an SMO algorithm for OC-SVM+ that significantly outperforms non-sequential algorithms for training the OC-SVM+ model. Its finite-time convergence is established. The experiments show how privileged information affects a descriptive domain in the space of original features. Comparative benchmark tests demonstrate that our algorithm is superior over interior point algorithms.

Problem

• One-Class SVM
• Privileged Information
• LUPI
• SMO Algorithm
• OC-SVM+

Innovation

• One-Class SVM+
• Sequential Minimal Optimization
• Learning Using Privileged Information
• Finite-Time Convergence
• Descriptive Domain Modeling