APAR pre-trains a tabular transformer on arithmetic combinations of target labels and fine-tunes it with adaptive feature masking, beating GBDT and neural baselines on 10 regression datasets.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
APAR: Modeling Irregular Target Functions in Tabular Regression via Arithmetic-Aware Pre-Training and Adaptive-Regularized Fine-Tuning
APAR pre-trains a tabular transformer on arithmetic combinations of target labels and fine-tunes it with adaptive feature masking, beating GBDT and neural baselines on 10 regression datasets.