Low-cost variable fidelity Bayesian support vector machine modeling of planar slot antennas

We demonstrate efficient Bayesian support vector regression (BSVR) modeling of coplanar waveguide-fed slot antennas with reduced training data sets. Coarse-discretization electromagnetic (EM) simulations are exploited in order to find a reduced number of training points used to establish a BSVR model of the antenna structure. Compared to conventional approximation-based models, the proposed technique allows substantial reduction (up to 48%) of the computational effort necessary to set up the models, with negligible loss in accuracy. Two antenna examples are demonstrated.