NOVA IMS and NOVA SBE join forces on the STEPS project, a DGES-funded initiative applying data science to tackle dropout rates and enhance student success for first-year undergraduates at Universidade NOVA de Lisboa.

STEPS adopts a multi-pronged approach:

• Unveiling Dropout Drivers: The project will identify key factors influencing student dropout and academic success at UNL through in-depth data analysis. This analysis will utilize advanced techniques like Decision Trees, k-nearest Neighbors, and Neural Networks.
• Predictive Modeling for Success: NOVA IMS will develop sophisticated predictive models to forecast dropout probabilities based on individual student data. This will allow for targeted interventions for at-risk students.
• Holistic Support Systems: The project will leverage insights from data analysis to design effective strategies for integrating new students, fostering innovative teaching practices, and establishing collaborative spaces for faculty.

By leading the data analysis and predictive modelling efforts, NOVA IMS will play a critical role in equipping NOVA with the tools to predict and mitigate student dropout. This initiative paves the way for evidence-based interventions to foster a thriving student body and achieve enhanced academic success.

Led by Professor Roberto Henriques, this project is integrated into the Deep Learning and Public Policy research lines within the MagIC research center, while also making significant contributions to the Education key impact area.