Steven Van Vaerenbergh

• Machine-learning techniques (kernel methods and online learning techniques):
  • nonlinear filtering in RKHS, kernel recursive least-squares (KRLS) etc.
  • kernel principal component analysis (kernel PCA) and spectral clustering
  • kernel canonical correlation analysis (CCA)
• ... and their application to signal processing and MIMO communication links:
  • identification and equalization of nonlinear systems
  • tracking with inertial measurement units (IMUs)
  • blind decoding of nonlinear or time-varying systems
  • blind source separation (BSS) and independent component analysis (ICA)

Downloads

• Connected Image Transformations algorithm: Matlab demo code.
• Kernel Recursive Least-Squares Tracker algorithm: code coming soon. Use the conference paper version in the meantime.
• Overlapping Mixtures of Gaussian Processes for the data association problem: Matlab source code.