A nonlinear regression model for estimating the size of web applications created using the laravel framework
Sergey Prikhodko, Natalia Prykhodko, Mikhail Vorona, Igor BelovolA three-factor nonlinear regression model for estimating the size of Web applications created using the Laravel framework is built on the basis of normalizing a four-dimensional non-Gaussian data set (actual size in thousands of lines of code; number of classes; the sum of the average number of classes affected by a given class and the average number of classes from which a given class receives effects; the average number of methods) using the multivariate Johnson transform for the SB family. The built model is compared with a linear regression model and nonlinear regression models based on the decimal logarithm and the univariate Johnson transform. The built model, in comparison with other regression models, has a smaller value of the average relative error and smaller widths of the prediction interval of nonlinear regression