Análise preditiva do desempenho acadêmico no Ensino Técnico Integrado ao Ensino Médio: aplicação de estatística, séries temporais e redes neurais em uma escola profissionalizante da rede federal

This research has analyzed the academic performance of first-year of technical integrated high school students (technical and int at a vocational school in the Federal Network, with an emphasis on the change in the assessment model that occurred between 2014 and 2017 and between 2018 and 2023. The study has compared the results of the assessments, which turned from four to three stages in order to assess the impact of this change on students` performance in basic cycle subjects. Quantitative data are used from regarding grades, school attendance and socioeconomic variables, such as family income and housing conditions. The statistical analysis shows that the reduction in the number of assessment stages did not harm student performance. A slight improvement in overall averages had been observed at the most recent period, indicating that the new format might favor more effective monitoring of learning. School attendance is shown to be a significant factor for academic success, highlighting the importance of encouraging student attendance. In addition, inequalities related to the socioeconomic context that influence academic performance were identified, reinforcing the need for educational policies aimed at promoting equity. In addition, predictive artificial intelligence models, including neural networks and time series, were applied to predict students' academic performance. These models performed some good performance, enabling the anticipation of results and the guidance of personalized pedagogical interventions. The integration of these technological tools has contributed to the improvement of the educational management process and to the development of more effective teaching strategies. It has been concluded that the changing the evaluation model to three stages has presented positive results for academic performance, without compromising the assessment of learning. The incorporation of advanced techniques on data has increased the potential for monitoring and intervention in the educational context.