Student Grade Pass/Fail Prediction: Secondary Education Performance in Portugal

Portuguese student data from two secondary schools, collected via school reports and questionnaires. The dataset contains attributes for grades, demographics, and social features, modeled for binary and multi-level classification tasks. It is based on the work of Cortez and Silva from 2008.

Use Cases

• Predicting final year grade (G3) pass/fail based on demographic and social features.
• Analyzing the correlation between period grades (G1, G2) and final performance.
• Identifying risk factors for student failure based on past class failures and study time.
• Modeling the impact of family educational support and internet access on academic achievement.

Strengths

• Includes multiple data sources: school reports and questionnaires.
• Contains a mix of numeric and boolean-converted demographic, social, and school-related features.
• Target attribute is defined for binary and five-level classification tasks.

Limitations

• Row count and file size are unknown, limiting suitability assessment.
• The dataset is a modified version lacking original attributes Mjob, Fjob, reason, and guardian.
• Last update date is unknown; freshness unverified.

Provenance

Source

Based on dipam7/student-grade-prediction, originally from Cortez and Silva, 2008.

Collection Method

Collected using school reports and questionnaires.

Geography

Portugal (two secondary schools: Gabriel Pereira and Mousinho da Silveira)