Machine learning techniques to estimate the degree of binder activity of reclaimed asphalt pavement

Abstract This paper describes the development of novel/state-of-art computational framework to accurately predict degree binder activity a reclaimed asphalt pavement sample as percentage indirect tensile strength (ITS) using reduced number input variables that are relatively easy obtain, namely compaction temperature, air voids and ITS. Different machine learning (ML) techniques were applied obtain most accurate data representation model. Specifically, three ML applied: 6th-degree multivariate polynomial regression with regularization, artificial neural network random forest regression. The produced models very similar precision, reporting mean absolute error ranging from 12.2 12.8% maximum ITS on test set. work presented in this is an evolution terms analysis results obtained within interlaboratory tests conducted by Task Group 5 RILEM Technical Committee 264 Reclaimed Asphalt Pavement. Hence, despite it has strong bonds framework, was developed independently can be considered natural follow-up.