Adaptive Weighted LSSVM for Multi-View Classification

TL;DR

AW-LSSVM introduces an adaptive, globally coupled LS-SVM framework for multi-view classification that enforces complementary learning by reweighting samples based on other views' misclassifications. The method iteratively solves per-view LS-SVMs with an additional error-weight term, updates weights via a decaying cross-view rule, and combines view-wise scores by averaging. Empirically, AW-LSSVM delivers consistent gains over kernel-based baselines on nine datasets, especially with many views, and shows privacy-friendly properties by keeping raw features local. The work positions AW-LSSVM as suitable for privacy-preserving settings such as vertical federated learning, with future directions including FL integration and semi-supervised extensions.

Abstract

Multi-view learning integrates diverse representations of the same instances to improve performance. Most existing kernel-based multi-view learning methods use fusion techniques without enforcing an explicit collaboration type across views or co-regularization which limits global collaboration. We propose AW-LSSVM, an adaptive weighted LS-SVM that promotes complementary learning by an iterative global coupling to make each view focus on hard samples of others from previous iterations. Experiments demonstrate that AW-LSSVM outperforms existing kernel-based multi-view methods on most datasets, while keeping raw features isolated, making it also suitable for privacy-preserving scenarios.