Linear variants of statistical data analysis

Voinov K. N., Nasr T. M., Afanasev M. Y., Hildayati A.

This paper is concerned with the formation of a bank of statistical data accumulated by the researcher when observing any physical phenomenon, for example, the wear and tear of parts in the operation of machinery, pressure, temperature, displacement, speed, the volume of gases emissions from the enterprises and from the exhaust pipes of cars and others. In this case, the simplest case corresponding to the linear dependence is initially considered, with a limited number of observation points of process development in time. Computer calculations using the mathematical shell of MathCad are shown. In addition, an algorithm for finding the equation determining the best approximating variant of the theoretical approximation of data. Moreover, an applied example is given with the use of a criterion by which it is possible to establish the following: Is it possible to leave in the general statistical summary of observations an unexpectedly appearing in the experiments sharply distinguishable value with respect to the others, which is apparently uncharacteristic, which will allow it then not to take into account.

Keywords:statistical data, wears, data bank, computer processing, information