PREDICT student PERFORMANCE before exam day.
Model the likelihood that a student will be a high performer based on study habits, lifestyle and environment — using a logistic regression model trained on thousands of real student records.
- 6,607
- Student records
- 98%
- Model accuracy
- 27
- Model features
- 67
- Median exam score
Performance predictor
Move the sliders and pick the options that best describe a student. The model updates instantly.
How it works
The app wraps a machine learning pipeline built with scikit-learn. Categorical inputs are encoded the same way the model was trained, so predictions stay consistent with the offline experiments.
Prepare the profile
Enter a student's academic, behavioural and environmental factors using the interactive form.
Run the model
Inputs are one-hot encoded and scored by a logistic regression model trained in Python.
Read the result
Get an instant prediction of whether the student is likely to be a high performer, with a confidence score.
Most influential factors
Ranked by the Random Forest feature-importance analysis.
- AttendanceConsistent class attendance
- Hours StudiedWeekly study time
- Previous ScoresPrior academic achievement
- Tutoring SessionsTargeted support
- Sleep HoursRest and recovery
About the target
The model predicts a High Performerlabel, which is derived from whether a student's exam score is above or below the median (67 points). Because the label is computed from the exam score, the very high accuracy partly reflects this relationship.
Logistic Regression reached 98.0% accuracy on the held-out test set, outperforming Random Forest (90%) and Decision Tree (83%).