In this paper, all of the discussed multivariate statistical methods
like MLR, NLR, and SLR have been utilized to construct the relevant
models. The predicted models are established based on explanatory
variable(s). The regression models are utilized to find out regression
coefficients. When fitting data with regression, our main objective is
often to discriminate between different models. The logical step following
model development is model validation. The purpose of verifi-
cation is to find the best model among three models. We can obtain
the relational expression among fitted freight volume, GDP and the
fixed investment based on MLR, NLR, and SLR: