Blast induced ground vibration which is one of the major aftermaths of blasting has over the years received wide attention in blasting studies due its adverse impact on the environment. In that regard, several studies have been conducted to accurately model and predict their level of occurrence. Thus, in this paper a support vector machine which is an artificial intelligence (AI) technique has been applied to model and predict ground vibration. In the development of the SVM model, a Bayesian optimisation algorithm was adopted to automatically select the hyperparameters. To ascertain its prediction performance, two empirical approaches namely: United State Bureau of Mines (USBM) and Indian standard models were also developed. The development of the various models was done using 210 blasting datasets from an open pit mine in Ghana. Performance of the various models were then evaluated using the mean absolute error (MAE), root mean squared error (RMSE) and correlation coefficient (R). Prediction results showed that the SVM model was able to accurately predict ground vibration compared to the empirical models by having the lowest training RMSE and MAE values of 0.1465 and 0.1199 respectively and highest training R value of 0.9027. Furthermore, it had the lowest testing RMSE and MAE values of 0.1533 and 0.1326 respectively and highest testing R value of 0.8476 and thus was ranked as the best model. This study has shown that the SVM model which is an AI technique is still a suitable tool for accurate modelling and prediction of blast-induced ground vibration.