The authors define an intervention-conditioned average treatment effect for event sequences and propose an IPW estimator with a Transformer-CNN intensity model, but the estimator's unbiasedness proof is flawed.
AdvancesinNeuralInformationProcessingSystems 34,17311–17324(2021)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Uncovering Causal Relation Shifts in Event Sequences under Out-of-Domain Interventions
The authors define an intervention-conditioned average treatment effect for event sequences and propose an IPW estimator with a Transformer-CNN intensity model, but the estimator's unbiasedness proof is flawed.