Causal relevance in nonlinear time-series models is better assessed via forecast necessity through edge ablation and prediction comparison than via coefficient magnitudes, as illustrated on democracy panel data.
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
-
Beyond Coefficients: Forecast-Necessity Testing for Interpretable Causal Discovery in Nonlinear Time-Series Models
Causal relevance in nonlinear time-series models is better assessed via forecast necessity through edge ablation and prediction comparison than via coefficient magnitudes, as illustrated on democracy panel data.