Abstract: Aiming at the time-consuming problem of full-wave simulation of the traditional graphene reconfigurable anantenna radiation characteristics, this paper uses Support Vector Regression (SVR) as a machine learning method for the rapid reconstruction prediction of graphene patch antenna parameters. The electromagnetic response of graphene patch antenna under different parameters (patch size, chemical potential, frequency, etc. ) is transformed into a regression estimation problem. Using the antenna element parameters as input and the corresponding S parameters as output, a regression model is established,and full-wave simulation software is used to establish a support vector regression training data set and test data set to achieve rapid prediction of the electromagnetic response of the graphene reconfigurable antenna element. In numerical examples, the prediction of S11 parameters is compared with the results of radial basis function network method and full-wave simulation software to verify the effectiveness of the proposed method.