Federated learning combined with pairwise-difference features ranks at-risk students from e-book logs as accurately as centralized training, with gains over non-differential baselines.
By employing a regression model, the extent of these differences can be explicitly learned, allowing the model to capture and utilize them
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Ranking-Based At-Risk Student Prediction Using Federated Learning and Differential Features
Federated learning combined with pairwise-difference features ranks at-risk students from e-book logs as accurately as centralized training, with gains over non-differential baselines.