A framework using structural causal models simulates parametric drifts to evaluate classifier robustness more realistically than static tests or noise perturbations.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Causal Parametric Drift Simulation: A Digital Twin Framework for Classifier Robustness Evaluation
A framework using structural causal models simulates parametric drifts to evaluate classifier robustness more realistically than static tests or noise perturbations.