摘要

The antenna design is a complicated and time-consuming procedure. This work explores using support vector machines (SVMs), a statistical learning theory based on the structural risk minimization principle and has a great generalization capability, as a fast and accurate tool in the antenna design. As examples, SVMs is used to design a rectangular patch antenna and a rectangular patch antenna array. Results show, after an appropriate training, SVMs is able to effectively design antennas with high accuracy.

全文