Student Performance Predictor
Individual project
A machine learning project involving student data analysis and academic performance prediction.
Overview
A machine learning project that analyses student data and predicts whether a student will pass or fail based on exam scores and background information.
Problem statement
Understanding which factors relate to academic performance can help identify students who may benefit from support. This project explores those relationships and whether a model can predict pass/fail outcomes from them.
Objectives
- Explore and visualise how student attributes relate to performance.
- Prepare the data for machine learning.
- Train a classification model to predict pass/fail outcomes.
- Evaluate the model with standard classification metrics.
Technologies
Data sources
- Students Performance in Exams (Kaggle) (opens in a new tab)
- Gender, race/ethnicity, parental level of education, lunch type and test preparation course
- Math, reading and writing scores
Methodology
- 01
Explore
Analysed the dataset with Pandas and visualised patterns with Seaborn and Matplotlib.
- 02
Prepare
Preprocessed the attributes for modelling.
- 03
Model
Trained a Random Forest classifier with Scikit-learn.
- 04
Evaluate
Assessed the model with accuracy, a confusion matrix and a classification report.
Screenshots
