In this paper, the generalized Radial Basis Function (RBF) neural network is applied to the short-term prediction of mine rockburst on a real-world dataset recorded by Huafeng Mine Company. For its network design and parameter learning, the Bayesian Ying-Yang (BYY) harmony learning algorithm and the synchronous LMS learning algorithm are utilized, respectively. It is demonstrated by the experimental results that this generalized RBF neural network based mine rockburst prediction method has obvious advantages of both prediction accuracy and convergence speed, and can satisfy the practical requirements of engineering application.