August 16, 2025

Constructing a binary prediction model with incomplete data: Variable selection to balance fairness and precision.

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Exploration of variable selection methods to enhance prediction accuracy and fairness in incomplete datasets.

Results reveal that BI-SS and SENET perform similarly in generalized linear models, but BI-SS excels in nested data designs.
Both methods utilize multiply imputed data sets effectively, yet they have different operational approaches to variable selection.
Bootstrap imputation-stability selection aggregates results across imputed datasets, enhancing accuracy and fairness.
This analysis highlights the importance of balancing fairness with precision when dealing with incomplete data.

He et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark