Data mining in educational filed becomes more involved and adding value to the educational research. In this study, the development of the professional identity of graduate students in a graduate school of science and engineering (GSSE) / Altinbas University were measured by professional identity-five factor scale (PIFFS). The results of the survey were analyzed statically using SPSS, as a result, four different levels of professional identity were recognized. The collected data analyzed by machine learning algorithms which were deployed using python code to predict student’s professional identity levels based on previously achieved results. Various types of algorithms were used and compared in a matter of accuracy and running time to select the most fitted model with the most accurate results.