ABSTRACT
The effective data processing algorithm is offered, which is the approximation of the plots of experimental points on the plane by piecewise linear functions. The idea of the method in the choice of the parameters of a linear function on a minimum sum of squared distances. Unlike the classical least squares method, this approach reduces the dispersion and leads to the dual selection of the parameters of the linear function. This algorithm is implemented as a program, the numerical experiments showed its high effectiveness.
