This study aims to explore and compare predictors for regular employment, income, or work satisfaction of 2 to 3-year college graduates and 4-year university graduates. To this end, the study applied the random forest method of the machine learning approaches to the 2018 Graduates Occupational Mobility Survey (2017GOMS) data collected by Korea Employment Information Service. According to the results of the study, 4-year university graduates had higher regular employment, income, and work satisfaction than those of 2 to 3-year college graduates. The random forest method showed that important variables affecting regular employment, income or satisfaction included ‘GPA,’ ‘the percentage of scholarship among tuition funding methods,’ ‘average working hours and income from in-school work experience,’ ‘reservation wage,’ and ‘the parents' average monthly income’ regardless of whether they are college or university graduates. Also, important variables selected for 4-year university graduates included ‘satisfaction with educational support facilities,’ ‘vocational education or training,’ and ‘school-to-work transition preparation after admission.’ On the other hand, important variables selected for 2 to 3-year college graduates included ‘day and night class,’ ‘major of study,’ and ‘importance attached when applying for a job (individual potential for development, employment stability, welfare benefits, social reputation, and commuting distance).’ Based on these results, implications and significance of this study were discussed.