A data science early-warning system for identifying academic risk among junior secondary students.
# Sokoto Student Success 🎓
## Student Academic Risk Early-Warning Dashboard
**Sokoto Student Success** is a Data Science capstone project demonstrating how
student learning indicators can be analysed and used to estimate academic-risk
categories for junior secondary students.
The project is designed around an education problem relevant to Sokoto:
teachers may benefit from identifying potential academic difficulties early,
before end-of-term results are available.
> **Important:** This project uses a fully synthetic dataset created for
> demonstration and learning. It contains no real student records and its
> findings must not be interpreted as empirical evidence about students in
> Sokoto State.
## Problem Statement
Students can experience declining academic performance for several reasons,
including attendance, previous assessment performance, assignment completion,
study habits and classroom participation. A teacher may not always have an
easy way to combine these indicators consistently.
This project explores whether a machine-learning workflow can provide an
early-warning signal that helps a teacher identify students who may require
additional monitoring or support.
## Objectives
1. Prepare and analyse a synthetic student-performance dataset.
2. Explore relationships among attendance, assessments, assignments,
homework, participation and study time.
3. Train and compare classification models.
4. Predict three academic-risk categories: High, Medium and Low.
5. Build an interactive dashboard for class-level analysis and individual
risk prediction.
6. Provide transparent, cautious explanations of model predictions.
7. Demonstrate an end-to-end Data Science workflow suitable for a portfolio
or training capstone.
## Dataset
The dataset contains **1,200 synthetic student records**.
Variables include:
- Age
- Gender
- Class level
- Attendance percentage
- Assignments given
- Assignments completed
- Previous test score
- Current CA score
- Class participat …