This study examines the problem of the success rate of computer course program participants at LKP Mitra Prestasi which is still an obstacle for institutional administrators. By applying a hybrid model approach in machine learning, which combines Random Forest and Long Short-Term Memory (LSTM) algorithms through the assembly merging technique, this study aims to develop a predictive model that is able to estimate the success rate of participants and recognize variables that contribute to the achievement of learning outcomes. The probability outputs from both models are used as an input feature for the Logistic Regression meta-classifier that studies the optimal combination of predictions to produce final graduation decisions. This approach leverages the Random Forest's advantages in recognizing static features and LSTM in recognizing temporal patterns. Research data was obtained from 100 course participants which included demographic information, academic performance, as well as data on the order of attendance and assignment submission. The research findings show that the hybrid model with the stacking ensemble built succeeded in achieving 95% accuracy, 95% precision, 100% recall, and 97.44% F1 score, with the main variables affecting graduation being final exam scores, attendance percentage, and accuracy of assignment submission.