Published in Scientific Papers. Series "Management, Economic Engineering in Agriculture and rural development", Vol. 15 ISSUE 4
Written by Agatha POPESCU

The goal of the paper was to chose the best regression model for milk production, the dependent variable, Y, and dairy bovine livestock, the independent variable, X, testing two polynomial regressions: linear regression and quadratic regression. The data were collected from the National Institute of Statistics for the period 2007-2014. The discrimination between models was based on the standard error of each type of regression, choosing the one which assured the highest accuracy of the prediction. As a conclusion, between milk production and dairy bovine livestock, it is a strong and positive correlation, r= 0.884, and about 78.30 % of milk production is determined by the milked livestock. As long as the Regression Standard Error was smaller in case of the Linear Regression, σest = 2,286.028830 than in case of the Quadratic Regression, σest = 2,336.915726, the linear regression having the formula y =23.974 x + 13,527 proved to better fit the data.

[Read full article] [Citation]

SPUPOPUPNO1