The article explores different ways to create accurate models for machining processes by analyzing force patterns and surface layer microgeometry. By comparing various methods of building regression models, researchers found that normalizing initial data leads to the smallest errors and most reliable models. Nonlinear relationships can only be found after statistical tests, and the "Italian cube" method of data normalization is crucial for accurate modeling. The F-criterion helps determine the adequacy of the models, with higher values indicating better reliability.