Effect of Data Imbalance on the Performance of Pavement Deterioration Models

Pavement performance models that can model consistently well across different pavement types, structures, traffic, and environmental conditions as throughout pavements’ lives are essential for the efficient management of transportation infrastructures. However, network data inherently imbalanced, most built on algorithms assume balanced distribution equal error cost. As a result, may vary groups. Considering this, objective this study was to evaluate imbalance its effect deterioration models. Specifically, analysis result longitudinal cracking is presented in paper. The mainly arises from two sources: intrinsic characteristics life expectancies. either design or process, whereas introduced by expectancies primarily process. used exhibits substantial skewness owing caused existence leads biased toward classes with more data.