Predicting student academic performance using TabPFN and SHAP: An interpretable machine learning approach

Qiang Yin · 2026

Read Predicting student academic performance using TabPFN and SHAP: An interpretable machine learning approach by Qiang Yin (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0328634.txt.

Description

Qiang Yin builds a system for predicting how high school students will perform, and for explaining its guesses, in PLOS ONE in 2026. Schools want early warning of struggling students, but many prediction models are black boxes and falter on modest datasets. Yin pairs a newer model called TabPFN, designed to learn from prior data patterns, with SHAP, a method that credits each input for its share of a prediction.

On records of four thousand Spanish high school students, the paired system beat seven rival models on accuracy. It also held up when the training data shrank: with only five hundred students it matched what older models needed thousands to achieve. The explanation step pointed to study habits, above all weekly study hours, attendance, and tutoring. Teachers get both a forecast and a reason they can act on.

Catalog record

CREATORQiang Yin
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE63739
LICENSECC_BY
RIGHTSCC BY
SOURCEPLOS

Source: https://doi.org/10.1371/journal.pone.0328634

Browse the full library · About the Tower of Babel