The infrastructure in Japan, which was mostly constructed in the 1960s, is facing aging problems. Higher priority has been given to the maintenance of infrastructure. The Road Committee of the Panel on Infrastructure Development recommended full-scale maintenance of aging roads in 2014. According to their recommendations, pavements may be inspected or replaced based on an appropriate renewal period depending on their deterioration level. This study aims to develop a numerical model to evaluate road surface irregularity using acceleration time histories recorded by smartphones. The records of the vertical component of acceleration of the automobile are employed to detect road sections with an international roughness index (IRI) of 12 mm/m or above. To achieve this objective, the IRI was measured with an interval of 10 m in the cities of Yokohama and Chofu. The authors employ logistic regression analysis and support vector machine to detect the road sections with an IRI of 12 mm/m or above. The discrimination accuracy is investigated, and a proper numerical model is proposed in this study.