Abstract:Recently, a new clustering method called maximum margin clustering (MMC) was proposed and has shown promising performances. It was originally formulated as a difficult nonconvex integer problem. To make the MMC problem practical, the researchers either relaxed the original MMC problem to inefficient convex optimization problems or reformulated it to nonconvex optimization problems, which sacrifice the convexity for efficiency. However, no approaches can both hold the convexity and be efficient. In this paper, a new linearithmic time sparse and convex MMC algorithm, called support-vector-regression-based MMC (SVR-MMC), is proposed. Generally, it first uses the SVR as the core of the MMC. Then, it is relaxed as a convex optimization problem, which is iteratively solved by the cutting-plane algorithm. Each cutting-plane subproblem is further decomposed to a serial supervised SVR problem by a new global extended-level method (GELM). Finally, each supervised SVR problem is solved in a linear time complexity by a new sparse-kernel SVR (SKSVR) algorithm. We further extend the SVR-MMC algorithm to the multiple-kernel clustering (MKC) problem and the multiclass MMC (M3C) problem, which are denoted as SVR-MKC and SVR-M3C, respectively. One key point of the algorithms is the utilization of the SVR. It can prevent the MMC and its extensions meeting an integer matrix programming problem. Another key point is the new SKSVR. It provides a linear time interface to the nonlinear kernel scenarios, so that the SVR-MMC and its extensions can keep a linearthmic time complexity in nonlinear kernel scenarios. Our experimental results on various real-world data sets demonstrate the effectiveness and the efficiency of the SVR-MMC and its two extensions. Moreover, the unsupervised application of the SVR-MKC to the voice activity detection (VAD) shows that the SVR-MKC can achieve good performances that are close to its supervised counterpart, meet the real-time demand of the VAD, and need no labeling for model training.

IMMC: Incremental Maximum Margin Criterion.

Image retrieval based on incremental subspace learning

Incremental semi-supervised subspace learning for image retrieval.

Incremental batch learning with support vector machines

WIMMC: Weighted Incremental Maximum Margin Criterion

An Incremental Updating Method for Support Vector Machines

AnIncremental Learning Algorithm Basedon Support Vector DomainClassifier

Incremental learning algorithm based on support vector machine

An Incremental Learning Algorithm Based on Support Vector Domain Classifier

Adaptive maximum margin criterion for image classification

Subspace maximum margin clustering.

Learning a Maximum Margin Subspace for Image Retrieval

Linear Time Maximum Margin Clustering

A scalable supervised algorithm for dimensionality reduction on streaming data

A Family of Maximum Margin Criterion for Adaptive Learning

Multiple Kernel Maximum Margin Criterion

Linearithmic Time Sparse and Convex Maximum Margin Clustering

SPARSE KERNEL MAXIMUM MARGIN CLUSTERING

Class Imbalance Robust Incremental LPSVM for Data Streams Learning.

Bayesian Maximum Margin Principal Component Analysis

Feature extraction via prototype margin distance maximizing criterion for subspace learning