Predictions using fuzzy regression models

Regression analysis is a powerful statistical tool used by engineers to predict infrastructure condition. The primary goal of the regression analysis is to find a best fitted mathematical model, so that a dependent variable can be forecasted from independent variables. Essentially, ordinary regression analysis is used to handle crisp data rather than fuzzy (non-crisp) data obtained from expert's subjective judgments. Therefore, the use of conventional regression technique is inappropriate when the dependent variable (condition rating) is a non-crisp or discrete variable. An An illustrative example regarding the cost estimate of excavation construction was exemplified. The results demonstrate the capability of the model that can assist cost engineers to better assess construction costs.