Codes

Smooth Support Vector Machine (SSVM) Toolbox

• Solve classification (SSVM) and regression (SSVR) problems

• Provide an automatic model selection for SSVM and SSVR with RBF kernels

• Can handle large scale problems by using reduced kernel (RSVM)

• Available at ssvm toolbox

Kernel Sliced Inverse Regression (SIR) Toolbox

• Including kernel PCA, kernel SIR, and kernel FDA.

• Associated SSVM(or SSVR) with KSIR in classification and regression problems.

• Available at KSIR toolbox

Robust Kernel Principal Component Analysis (PCA) Package

• Extract kernel principal components via psi-principle

• Associated with tuning procedure

• Available at robust KPCA toolbox

Anomaly Detection via Over-sampling Principal Component Analysis

• Utilize the variation of principal directions to detect outliers

• Apply the power method for exact over-sampling PCA (osPCA)

• Apply recursive least squares for online osPCA

• Available at osPCA

Heterogeneous Domain Adaptation and Classification by Exploiting the Correlation Subspace

• Reduced kernel Canonical Correlation Analysis (CCA)

• Correlation-Transfer Support Vector machine

• Available at CCATF