Targeted Maximum Likelihood Learning

By M. J. van der Laan, Donald B. Rubin. 2006

1,011 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Statistics and Probability, Statistical Learning Theory

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistics and Probability

In Statistical Learning Theory

Citations in the corpus

Machine Learning Methods That Economists Should Know About (2019, methodology): “Targeted maximum likelihood [van der Laan and Rubin, 2006] is one approach to this, while Imai et al. [2013] proposed using LASSO to uncover heterogeneous treatment effects.”
Double/De-Biased Machine Learning of Global and Local Parameters Using Regularized Riesz Representers (2018, background): “We use notation from the empirical process theory, see Van Der Vaart and Wellner (1996).”
Double/Debiased/Neyman Machine Learning of Treatment Effects (2017, methodology): “Moreover, the use of efficient scores could be considerably refined using the targeted maximum likelihood approach of Scharfstein, Rotnitzky, and Robins (1999) and van der Laan and Rubin (2006) in many contexts.”
Approximate residual balancing: debiased inference of average treatment effects in high dimensions (2016, methodology): “Xi/ obtained by elastic net logistic regression, with the propensity scores trimmed at 0:05 and 0:95; AIPW , which pairs elastic net regression adjustments with the above inverse propensity weights (9); the weighted elastic net, motivated by Kang and Schafer (2007), that uses inverse propensity weig”
Automated versus Do-It-Yourself Methods for Causal Inference: Lessons Learned from a Data Analysis Competition (2017, positive): “The propensity score is also an important component of the Targeted Maximum Likelihood Estimation (TMLE) (van der Laan and Robins, 2003; van der Laan and Rubin, 2006) approach used in several of the submitted methods.”
Faster Rates for Policy Learning (2017, methodology): “(17) Observe that above Ḟπ(V(π)) only enters fπ as a multiplicative constant, and therefore an estimating equation-based estimator or TMLE for V(π) can be asymptotically linear for V(π) without estimating Ḟπ.”
A Survey of Learning Causality with Data (2018, methodology): “core models is developed for the factual outcomes. Then we can estimate ATE by taking the average over the estimated ITE for all the instances as in Eq.9. Targeted Maximum Likelihood Estimator (TMLE) [154] is a more generalized method than DRE. ATE can be inferred with TMLE as: 1 n ˝ n i=1 Q ∗(1,x i”
TREE-BASED REINFORCEMENT LEARNING FOR ESTIMATING OPTIMAL DYNAMIC TREATMENT REGIMES (2018, neutral): “…[Murphy, van der Laan and Robins (2001), Wang et al. (2012), Hernán, Brumback and Robins (2001)], G-estimation of structural nested mean models [Robins (1994, 1997, 2004)], targeted maximum likelihood estimators [van der Laan and Rubin (2006)] and likelihood-based approaches [Thall et al. (2007)].”
The Impact of Machine Learning on Economics (2018, methodology): “See Varian (2014) for an overview of some of the most popular methods and Mullainathan and Spiess (2017) for more details.”
Causal machine learning for predicting treatment outcomes (2024, neutral): “23,27,49,53,54,61,98,99).”
Augmented minimax linear estimation (2017, neutral): “…m̂(·) and γ̂ψ(·) for the regression function and the Riesz representer, and then plug them into (Chernozhukov et al., 2016; Newey, 1994; Robins and Rotnitzky, 1995) ψ̂DR = γ(m̂)− 1 n n∑ i=1 γ̂ψ(Zi) (m̂(Zi)− Yi) (11) or an asymptotically equivalent expression (see e.g. van der Laan and Rubin, 2006).”
Targeted Undersmoothing. (2017, neutral): “24See, for example, [15], [52], [43], [51] for classic examples.”
Causal inference and counterfactual prediction in machine learning for actionable healthcare (2020, background): “We wish to generalize if the treatment works also in the population of city B (target environment) where only observational data is available, since it happens that the age distribution in city A, P ( z ), is different than that P * ( z ) in city B.”