NihmaMT/student-scholarship-predictor

Student Scholarship Predictor

A machine learning web app that predicts whether a student is likely to receive a scholarship based on their academic performance and attendance.

Dataset

The app uses Student_performance_dataset.csv, which contains records for ~6,000 students with the following fields:

• —Student_ID / Student_Name — identifiers (not used for prediction)
• —Math_Score, Science_Score, English_Score — subject scores (0–100)
• —Attendance_Percentage — percentage of classes attended
• —Age — student age (17–30)
• —Gender — Male, Female, or Other
• —Department — academic department (CS, IT, ECE, MECH, CIVIL)
• —Scholarship — target label: Yes or No

One row with a missing Scholarship value is dropped at startup.

Model

A Random Forest Classifier is trained with:

• —80/20 stratified train/test split (random_state=42)
• —n_estimators=200, max_depth=5, min_samples_leaf=10
• —class_weight="balanced" to handle any label imbalance
• —Label encoding for Gender, Department, and Scholarship

The live test accuracy is displayed in the app header.

Usage

Adjust the sliders and selectors to match a student's profile, then click Predict to see the scholarship outcome and confidence breakdown.