This document discusses a study that uses regression analysis techniques to estimate surface roughness values based on tribological parameters during hard turning of AISI 52100 steel. Experiments were conducted using a Taguchi L9 orthogonal array with cutting speed, feed rate, and depth of cut as parameters. Linear, quadratic, and cubic regression models were developed to correlate the parameters to surface roughness. The results showed that the estimated correlations predicted roughness with 99.5-99.9% accuracy, with only 6.19% uncertainty in the experiments.