On a synthetic causal model, a multilayer perceptron with the Mish activation function predicts the lower and upper bounds of the probability of necessity and sufficiency for subpopulations with limited data, reaching a mean absolute error of about 0.02.
An axiomatic characterization of causal counterfactuals
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Estimating Probabilities of Causation with Machine Learning Models
On a synthetic causal model, a multilayer perceptron with the Mish activation function predicts the lower and upper bounds of the probability of necessity and sufficiency for subpopulations with limited data, reaching a mean absolute error of about 0.02.