Predicting Resonant frequency of Annular ring Compact microstrip antenna using various Machine Learning Techniques

Three powerful prediction machine learning (ML) methods Support vector machine (SVM), Gaussian Process Regression (GPR) and Artificial neural network (ANN) are used to compute the resonant frequency of annular ring compact microstrip antenna (ARCMA) for dominant mode T M₁₁ is presented. The designed antenna can work in the band of 0.75336 GHz to 5.7628 GHz for the Ultra high frequency (UHF) band and Super high frequency (SHF) band applications. For preparing model, a total 100 data samples were collected from varying the ratio of an outer radius to the inner radius of ARCMA using CST^TM, and 90 data samples were used to build three models, and 10 data samples used for testing these models. Based on prediction accuracy for new data, it is found that GPR model gives a more accurate result. The predicted outcome shows that SVM, GPR, and ANN methods can be applied to compute the resonant frequency of ARCMAs accurately.