 This paper proposes a novel approach to designing a reconfigurable multiband graphene patch antenna for use in the Terahertz frequency range. A deep learning algorithm is used to accurately predict the antenna's performance parameters, such as gain, beam steering, and return loss. The predicted parameters are then used to design five band and three band antennas which can be tuned to operate within the desired frequency ranges. This article was authored by Mohamed Mashiqi, Poryekabiri, Amir Samand Noamin, and others.