Document Type : Primary Research paper
National Technological University of South Lima, Lima, Perú
Private University of the North, Lima, Perú
National University of Huancavelica, Huancavelica, Perú
Academic performance is related to academic success, postponement and desertion, the latter two are problems that have increased in Peru due to the history of the pandemic, which has made it notorious. The digital transformation offers educational institutions opportunities to integrate their members with technological and cultural changes through participation in accordance with their educational role, and become the engine of educational reforms, in this case, the technological system that constitutes the ecosystem. Data forms information, knowledge and actions. In this sense, the use of data mining methods such as CRISP-DM and the application of machine learning algorithms provide the opportunity to design adjustable work models according to each institution to predict academic performance and determine the reasons for delays and dropouts. academics. Therefore, by applying a development and feedback cycle, you can improve the certainty of the university's intellectual capital and machine learning technology to optimize academic predictions and contribute to the development of students, society and institutions of social education. Finally, we conclude that the academic performance of universities can be satisfactorily predicted, and machine learning methods are being implemented in different regions of the world to improve intellectual capital and institutional performance.