In the present study, the prediction of cutting forces and surface roughness was carried out using neural networks and support vector regression (SVR) with six inputs, namely, three axis vibrations of the tool holder and cutting speed, feedrate and depth of cut. The data obtained by experimentation are used to construct predictive models. A feedforward backpropagation neural network and SVR have been selected for modelling. The coefficient of determination (R 2), mean absolute prediction error and root mean square error were calculated for each method, and these values served as a measure of prediction precision. We carried out comparison of the prediction accuracy of artificial neural networks and SVR. Comparison of the two models indicates that both models have successful performance. Experimental results are provided to confirm the effectiveness of this approach. © 2012 Institute of Materials, Minerals and Mining.