Information Technology Journal1812-56381812-5646Asian Network for Scientific Information10.3923/itj.2009.615.618XiangyangMu TaiyiZhang 4200984This study presents an empirical study for Minimax Probability Machines (MPM) for prediction. Considering that the Euclidean distance has a natural generalization in form of the Minkovsky’s distance, a novel MPM using Minkovsky’s norm in Gaussian kernel function is proposed. The performance of proposed method is evaluated with the prediction for Ethernet traffic data. Result shown that the novel MPM here in using Gaussian kernels with Minkovsky’s distance (α=1) and (α=5) can achieve better prediction accuracy than the Euclidean distance.]]>Aronszajn, N.,1950Burges, C.J.C.,1998Chen, B.S., S.C. Peng and K.C. Wang,2000Huang, K., Y. Haiqin, K. Irwin, R.L. Michael and C. Laiwan,2004Lanckriet, G.R.G., L.E. Ghaoui, C. Bhattacharyya and M.I. Jordan,2002Lanckriet, G.R.G., L.E. Ghaoui, C. Bhattacharyya and M.I. Jordan,2002Liu, Z.X., D.Y. Zhang and H.C. Liao,2005Perez-Cruz, F. and O. Bousquet, 2004Ribeiro, B.,2002