JFET_V3_N2_RP2 P. Nanda Kumar G. Ranga Janardhana Kim S. Kun Journal on Future Engineering and Technology 2230 – 7184 3 2 7 15 Surface roughness, Factorial design of experiments, Prediction models, Artificial neural network In the present day, the important goal in the modern industries is to manufacture high quality and low cost products in just in time. The quality of the products depends on the surface roughness and hence the surface roughness plays an important role in product manufacturing. In this paper, AI based neural network modeling approach is presented for the prediction of surface roughness of Aluminum Alloy products machined on CNC Turning Center. The experiments were conducted based on the principle of Factorial Design of Experiment (DOE) method. Trails were made with different combinations of step size and momentum to select the best learning parameter. The best network structure with least MSE was selected among the several networks. The multiple regression models, which are most widely used as prediction methods, are considered to compare the developed ANN model performance. November 2007 - January 2008 Copyright © 2008 i-manager publications. All rights reserved. i-manager Publications http://www.imanagerpublications.com/Article.aspx?ArticleId=658