Predicting the use of artificial intelligence based on digital competencies in postgraduate education using a machine learning approach

Introduction – The accelerated development of artificial intelligence and machine learning has transformed postgraduate education by incorporating predictive analytics and data-driven decision-making processes. In this context, digital competencies are a key factor for the effective use of artificial intelligence tools in academic environments. Therefore, this study aims to determine to what extent machine learning models can predict the use of artificial intelligence based on digital competencies among postgraduate students at the State University of Milagro. Methods – A quantitative, applied approach was adopted, with a non-experimental, cross-sectional design at an explanatory–predictive level. The sample consisted of 261 postgraduate students selected through non-probabilistic convenience sampling, based on accessibility and voluntary participation criteria. Data were collected using a structured Likert-type questionnaire and analyzed through statistical techniques in SPSS and predictive modeling using machine learning algorithms, including logistic regression and random forest. Model performance was evaluated using accuracy, AUC, precision, recall, and F1-score. Results – The findings revealed high predictive performance of the models, with AUC values above 0.90, highlighting logistic regression and random forest as the most stable algorithms. Instrumental digital competencies were identified as the main predictor of artificial intelligence use, surpassing critical, informational, and sociodemographic variables. Conclusion – The use of artificial intelligence in postgraduate education can be significantly predicted based on the level of digital competencies, confirming a functional relationship between both variables and emphasizing the need to strengthen these competencies to support effective integration of artificial intelligence in higher education.