Causality and Generalizability: Identifiability and Learning Methods

TL;DR

This thesis advances causal inference by developing novel estimators, exploring distributional robustness, and proposing new structure learning methods for causal models, with theoretical guarantees and practical implications.

Contribution

It introduces new consistent causal effect estimators, links distributionally robust prediction to econometrics, and proposes a structure learning method for additive noise models with directed trees.

Findings

Proposed estimators outperform canonical methods in certain settings.

Connected distributionally robust prediction to econometric estimators.

Proved consistency and error control for the new structure learning method.

Abstract

This PhD thesis contains several contributions to the field of statistical causal modeling. Statistical causal models are statistical models embedded with causal assumptions that allow for the inference and reasoning about the behavior of stochastic systems affected by external manipulation (interventions). This thesis contributes to the research areas concerning the estimation of causal effects, causal structure learning, and distributionally robust (out-of-distribution generalizing) prediction methods. We present novel and consistent linear and non-linear causal effects estimators in instrumental variable settings that employ data-dependent mean squared prediction error regularization. Our proposed estimators show, in certain settings, mean squared error improvements compared to both canonical and state-of-the-art estimators. We show that recent research on distributionally robust prediction methods has connections to well-studied estimators from econometrics. This connection leads us to prove that general K-class estimators possess distributional robustness properties. We, furthermore, propose a general framework for distributional robustness with respect to intervention-induced distributions. In this framework, we derive sufficient conditions for the identifiability of distributionally robust prediction methods and present impossibility results that show the necessity of several of these conditions. We present a new structure learning method applicable in additive noise models with directed trees as causal graphs. We prove consistency in a vanishing identifiability setup and provide a method for testing substructure hypotheses with asymptotic family-wise error control that remains valid post-selection. Finally, we present heuristic ideas for learning summary graphs of nonlinear time-series models.