Approximation efficiency for multivariable dependencies in metal cutting

Problems to improve mathematical and software means are analysed using approximation of the functions by multivariable polynomial models. Experimental dependencies of the machining tool life are presented as examples. Stochastic approximation method is applied as an algorithm of deriving the coefficients. Analysis of the presented method shows that the approximation error tends close to the experimental error. This method is recommended to approximate the functions of several variables at making various experiments.