Análise de Classificadores para Predição de Evasão em um Curso Superior de Tecnologia da Informação de um Instituto Federal

School dropout has been a long-standing issue in Higher EducationInstitutions (HEIs), raising concerns and prompting mitigationefforts. This paper examines dropout in a Technology course ata Federal Institute, addressing the research question: "How canmachine learning classifiers assist in predicting student dropout?".The study applies machine learning classifiers using data from aFederal Institute, including demographic (age group, income, city,special needs, ethnicity, gender) and academic variables (admissionmethod, vacancy type, enrollment status). The objective is todevelop analyses ranging from descriptive statistics to predictivemodeling. The classifiers used include Support Vector Machine, RandomForest, Logistic Regression, Gradient Boosting, AdaBoost, andK-Nearest Neighbors. The goal is to support teachers, coordinators,and administrators in implementing preventive measures such aspersonalized mentoring, continuous monitoring, and institutionalpolicies to improve academic infrastructure and inclusivity. Theresults show that Support Vector Machine, Gradient Boosting, andAdaBoost achieved the best performance, with F1-measure valuesbetween 0.6 and 0.8, demonstrating their predictive capability. Thishighlights the potential of machine learning in addressing studentdropout in higher education.