| Popis: |
The state of the art currently for addressing pavement deterioration proposes the development of Pavement Deterioration Models, using a clusterwise approach that requires a priori knowledge of the optimal number of clusters as well as significant explanatory variables. In addition, the objective function used to solve the clusterwise problem is the minimization of the sum of squared errors, which always decreases with additional cluster(s) and/or explanatory variable(s). To address these limitations, a mathematical programming framework is proposed based on the Bayesian Information Criterion, which does not require a priori information about the optimal number of clusters. An extensive optimization approach was used to find a solution to the proposed mathematical program, and issues associated with overfitting were investigated. Results using data from the entire State of Nevada illustrate the advantage of the proposed framework. |
