PuckTrick library adds controlled imperfections to synthetic data and shows that models trained on the resulting contaminated data outperform those trained on clean synthetic data in financial dataset experiments.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
PuckTrick: A Library for Making Synthetic Data More Realistic
PuckTrick library adds controlled imperfections to synthetic data and shows that models trained on the resulting contaminated data outperform those trained on clean synthetic data in financial dataset experiments.