Formalizes the max-potency problem for cost-constrained experiment selection to tighten bounds on causal queries, proves NP-hardness, and introduces two graphical pruning criteria that reduce the search space by 50-88% on average.
Embedding hard physical constraints in neural network coarse-graining of 3d turbulence.arXiv preprint arXiv:2002.00021, 2020
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
verdicts
UNVERDICTED 4representative citing papers
Parametric models for principal causal effects produce only partial identification without principal ignorability, with association parameters for strata identifiable solely under violation of that assumption plus strong parametric constraints.
Decomposing predictions into between-unit, within-unit-across-time, and counterfactual components shows within-unit accuracy is a structurally better proxy than overall accuracy for recovering true causal treatment effects from non-experimental panel data.
A single-layer architecture called FlowMixer uses constrained matrix operations and a semi-group property to enable depth-agnostic, interpretable spatiotemporal forecasting with direct eigenmode extraction.
citing papers explorer
-
Optimal Experiments for Partial Causal Effect Identification
Formalizes the max-potency problem for cost-constrained experiment selection to tighten bounds on causal queries, proves NP-hardness, and introduces two graphical pruning criteria that reduce the search space by 50-88% on average.
-
Partial identification of principal causal effects under violations of principal ignorability
Parametric models for principal causal effects produce only partial identification without principal ignorability, with association parameters for strata identifiable solely under violation of that assumption plus strong parametric constraints.
-
Prediction decomposition for causal analysis
Decomposing predictions into between-unit, within-unit-across-time, and counterfactual components shows within-unit accuracy is a structurally better proxy than overall accuracy for recovering true causal treatment effects from non-experimental panel data.
-
FlowMixer: A Depth-Agnostic Neural Architecture for Interpretable Spatiotemporal Forecasting
A single-layer architecture called FlowMixer uses constrained matrix operations and a semi-group property to enable depth-agnostic, interpretable spatiotemporal forecasting with direct eigenmode extraction.