Establishment of correlation model between compositions and dynamic viscoelastic properties of asphalt binder based on machine learning

•Datasets of macro and micro characteristic parameters of asphalt binders were established.•The absorption peak at 1700 cm−1 was the most sensitive to oil sources.•The composition parameters A1700, A870, A810, A745, A720, I1600, N, S, C/N were chosen as input variables to predict viscoelastic properties.•Support vector machine (SVM) was an effective method for predicting the viscoelastic properties of asphalt binders.