Home / Work / Student Grades

Using Linear Regression to Predict Student Grades

Linear regression project
Problem

Can a student’s final grade be predicted from behavioural and historical data — study time, previous results, failures, and absences?

Approach
  • Explored and preprocessed a student performance dataset
  • Selected relevant features and split train/test data
  • Trained a linear regression model and evaluated accuracy
Result

A working predictive model demonstrating how interpretable regression can surface which factors most influence academic outcomes — a foundation for more complex modelling approaches.

Stack
  • Python
  • Scikit-learn
  • Pandas

Features used

The model incorporated study time, previous academic results, number of past failures, and absences — giving a multidimensional view of factors that correlate with final performance.

View on GitHub ← All work