Modeling of a High-Speed Link Based on an Efficient Implementation of Vector-Valued Machine Learning Kernel Regression

This paper presents an efficient implementa- tion of the vector-output kernel Ridge regression (KRR). The proposed approach is applied to model the frequency- domain behavior of the magnitude of the transfer function of a parametric high-speed link as a function of 11 parameters. The accuracy and the computational cost of the proposed technique are assessed on noisy samples and compared with the ones of a state-of-the-art modeling technique based on the combination of the principal components analysis (PCA) and the least-squares support vector machine (LS-SVM) regression.