A preprocessing framework that reallocates transformed-space width to minimize the target function's Dirichlet energy shows aggregate gains on tabular benchmarks, with the supervised variant best in regression.
An inductive bias for tabular deep learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Tabular Numeric Stretch Transformation
A preprocessing framework that reallocates transformed-space width to minimize the target function's Dirichlet energy shows aggregate gains on tabular benchmarks, with the supervised variant best in regression.