Key Insights: Built a data analytics solution to identify risk factors and predict academic outcomes for high school students, helping address dropout rates that can reach over 40% in developing countries.

About the Client

In this case, the client was a national government agency responsible for secondary education nationwide, alongside the National Institute for Educational Evaluation.

The Challenge

Student dropout is a major challenge for education systems, with rates ranging from 5% to over 40% in developing countries. Early identification of students at risk is critical to designing effective prevention strategies.

The goal was to prevent school dropout by detecting which students were at risk of leaving school early. The project aimed to study different student trajectories and identify risk factors based on socio-economic and academic data of secondary school students in Uruguay, using datasets of more than 100,000 students from across the country.

Marvik’s Approach

We developed a data analytics solution that:

• Identified vulnerability factors linked to higher dropout risk.
  
• Performed clustering to group students by educational trajectories.
  
• Generated predictions of a student’s academic outcome for a given year.
  

The Technology

The solution was built in Python, leveraging TensorFlow and PyTorch for modeling, alongside advanced data analysis, clustering, and visualization techniques to process and extract insights from a national dataset of over 100,000 students.

The Results & Impact

• Provided actionable insights to guide intervention strategies.
  
• Allowed early identification of at-risk students across a national dataset.
  
• Gave policymakers and educators the tools to better allocate resources and improve retention rates.
  

With dropouts impacting long-term social and economic outcomes, this approach empowers education systems to intervene before it’s too late.

Why This Matters

By applying analytics at scale, this project helps shift education systems from reactive to proactive, improving the chances of success for thousands of students.