Input variables & classification
In addition to the inputs tested in the previous studies (Kim et al.
2011; Kim, Wunneburger, & Neuman, 2013), more variables are
added to test the robustness of the proposed SDSS. Table 1 summarizes
the selected input variables and they are in five categories.
In this study, a 1-to-5 scale is utilized although a 1-to-9 scale is
generally recommended by the previous researchers (Gorsevski
et al. 2013; Saaty, 2013), because using a 5 scale increases the
sensitivity of variable effects. As mentioned earlier, a 1-to-9 scale
with 15 different variables creates minimal changes as the variances
required in this case is too narrowed. In addition, a reversed
scale is utilized, a value of ‘5’ reflects the least desirable value for
locating a transportation corridor in that place, and a value of ‘1’
reflects a suitable value for locating a transportation corridor. It
means the lower the suitability scores, the more desirable for a
transport corridor.