Application of Modern Regression Techniques and Artificial Neural Networks to Pavement Prediction Modeling

This study strives to illustrate the benefits of incorporating the principles of dimensional analysis, subject-related knowledge, and statistical knowledge into pavement prediction modeling process. Modern regression techniques including local regression and regression splines as well as back propagation neural networks were briefly introduced. Factorial 2-D and 3-D finite element runs and BISAR runs for different pavement systems were conducted to generate the deflection databases for the analysis. The resulting ANN model using all dominating dimensionless parameters was proved to have higher accuracy and require less network training time than the other counterpart using purely input parameters. Increasing the complexity of ANN models does not necessarily improve the modeling statistics. The results also showed that using higher number of neurons and hidden layers sometimes lead to even worse modeling statistics which was an indication of over training and should be avoided. Several local regression models requiring minimal amount of modeling time were also developed using the same databases.