School dropout is a critical problem that negatively impacts the social and economic development of many countries, including Peru. Despite the efforts of the main government institutions, the lack of effective predictive tools hinders the early identification of the most at-risk students. This study proposes a classification model based on machine learning techniques to predict school dropout from sociodemographic and academic data. A five-phase methodology was applied: Data collection; Pre-processing; Feature selection (Logistic Regression, XGBoost, Random Forest, and SelectBest); Model implementation (Random Forest, SVM, XGBoost, Decision Tree, KNN, Naive Bayes, and Logistic Regression); and Evaluation. The results showed that the best performing model was Logistic Regression with SelectKBest feature selection reaching 93.07% in Accuracy, 90% in Recall and 80% in F1-Score. In conclusion, the results demonstrate that it is possible to build intelligent systems capable of classifying school dropouts, which will enable the responsible entities to implement more effective preventive strategies.