Using Machine Learning Algorithms to Predict the Training Needs of Students for SMK Pustek Tangerang

Students at SMK Pustek Serpong in South Tangerang have diverse backgrounds, interests, and potentials that need to be identified and developed through appropriate training programs. This research aims to utilize machine learning algorithms to improve the accuracy of predicting students' training and development needs. Student data, including demographics, academic achievements, interests, and extracurricular activities, will be used to train models such as Random Forest Classifier, SVM, Gradient Boosting Classifier, and K-NN, targeting their chosen academic majors. The problem-solving approach involves problem identification, selection of machine learning methods, dataset collection, and model implementation. The research findings show that Gradient Boosting Classifier performs best with 77% accuracy, 79% precision, 96% recall, and an F1-score of 87% for the majority class. Conversely, K-NN achieves 67.97% accuracy but exhibits lower performance in identifying minority classes with precision and recall around 28% and 23%, respectively.