Student performance: prediction & dashboard

Problem

Support works best when it arrives early. This project predicts student academic outcomes from behavioural and demographic data, so educators can identify students who may be at risk and act with timely intervention rather than after the fact.

Approach

It pairs a Machine Learning (ML) prediction model with an interactive dashboard. The model learns from historical student data; the dashboard lets users explore that data visually and see predicted performance as they adjust inputs.

Methods & models

• Classification and regression models trained on student features
• An interactive Streamlit app for exploration and live prediction

Why it matters

The aim is to support judgement, not replace it — giving educators an early signal they can combine with everything else they know about a student.

Tools & technologies

Code & links

• Live dashboard — explore the data and run predictions.
• Dashboard repository
• Model repository