Demystifying statistical learning based on efficient influence functions

Abstract

Evaluation of treatment effects and more general estimands is typically achieved via parametric modelling, which is unsatisfactory since model misspecification is likely. Data-adaptive model building (e.g. statistical/machine learning) is commonly employed to reduce the risk of misspecification. Naive use of such methods, however, delivers estimators whose bias may shrink too slowly with sample size for inferential methods to perform well, including those based on the bootstrap. Bias arises because standard data-adaptive methods are tuned towards minimal prediction error as opposed to e.g. minimal MSE in the estimator. This may cause excess variability that is difficult to acknowledge, due to the complexity of such strategies. Building on results from non-parametric statistics, targeted learning and debiased machine learning overcome these problems by constructing estimators using the estimand's efficient influence function under the non-parametric model. These increasingly popular methodologies typically assume that the efficient influence function is given, or that the reader is familiar with its derivation. In this paper, we focus on derivation of the efficient influence function and explain how it may be used to construct statistical/machine-learning-based estimators. We discuss the requisite conditions for these estimators to perform well and use diverse examples to convey the broad applicability of the theory.

References in corpus (5)

• Higher order influence functions and minimax estimation of nonlinear functionals
• Adaptive nonparametric confidence sets
• DoubleML -- An Object-Oriented Implementation of Double Machine Learning in Python
• Estimating Structural Target Functions using Machine Learning and Influence Functions
• Tutorial: Deriving The Efficient Influence Curve for Large Models

Cited by in corpus (17)

• Scientific Inference With Interpretable Machine Learning: Analyzing Models to Learn About Real-World Phenomena
• Model-robust and efficient covariate adjustment for cluster-randomized experiments
• A Flexible Approach for Predictive Biomarker Discovery
• Normalized Augmented Inverse Probability Weighting with Neural Network Predictions
• A nonparametric framework for treatment effect modifier discovery in high dimensions
• Variable importance measures for heterogeneous treatment effects
• Incremental Propensity Score Effects for Criminology: An Application Assessing the Relationship Between Homelessness, Behavioral Health Problems, and Recidivism
• Recovery and inference of causal effects with sequential adjustment for confounding and attrition
• Using Individualized Treatment Effects to Assess Treatment Effect Heterogeneity
• Doubly robust estimation of optimal treatment regimes for survival data using an instrumental variable
• Revisiting the propensity score's central role: Towards bridging balance and efficiency in the era of causal machine learning
• Parameterising the effect of a continuous treatment using average derivative effects
• Valid causal inference with unobserved confounding in high-dimensional settings
• Continuous difference-in-differences with double/debiased machine learning
• Decorrelated Variable Importance
• Efficient nonparametric estimators of discrimination measures with censored survival data
• A framework for joint assessment of a terminal event and a score existing only in the absence of the terminal event