In this paper, the damage state classification method based on machine learning is applied to a shaking table test data of a full-scale three-story steel frame. The following results are shown; 1) By machine learning using only the shaking table tests data, it is possible to estimate each damage state classification by linear discriminant analysis. 2) Acceleration data from response analysis as the training data and shaking table test data as the verification data are utilized for machine learning, and damage state classification can be estimated.