Optimal Structural Nested Models for Optimal Sequential Decisions

By James M. Robins. 2004

660 citations recorded by the source corpus.

How researchers see this paper

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

In Statistics and Probability

Citations in the corpus

Learning Optimal Fair Policies (2018, methodology): “Consistent estimators of ψ can be obtained using the estimating equations E [ U ( ψ, ζ ( ψ ) , α )] = 0 , as shown in (Robins 2004).”
Deep Reinforcement Learning for Dynamic Treatment Regimes on Medical Registry Data (2017, background): “Methods to identify the optimal DTRs have recently received attentions in the statistical community [7], [8], [9], [10], [11], [12], [13], [14], and most of exsiting methods focused on data from randomized clinical trials.”
Learning When-to-Treat Policies (2019, methodology): “To identify causal effects, we make the standard assumptions of sequential ignorability, consistency and overlap (Robins 1986, 2004; Hernán, Brumback, and Robins 2001; Murphy 2003).”
Reinforcement Learning Strategies for Clinical Trials in Non-small Cell Lung Cancer (2011, background): “The concept is an extension and melding of dynamic treatment regimes in counterfactual frameworks (Murphy, 2003;Robins, 2004) and sequential multiple assignment randomized trials (Murphy, 2005a) to accommodate an irreversible disease state with a possible continuum of treatment options.”
Quasi-oracle estimation of heterogeneous treatment effects (2017, background): “…inference and constructions of orthogonal moments (Athey and Wager, 2017; Belloni et al., 2017; Bickel et al., 1998; Luedtke and van der Laan, 2016; Newey, 1994; Robins, 2004; Robins and Rotnitzky, 1995; Robins et al., 2017; Scharfstein et al., 1999; Tsiatis, 2007; van der Laan and Rose, 2011).”
Q-learning Residual Analysis: Application to The Effectiveness of Sequences of Antipsychotic Medications for Patients with Schizophrenia (2016, neutral): “The authors of [15] consider model checking with residual diagnostic plots for an alternative estimation method called g-estimation [16].”
Optimal a priori balance in the design of controlled experiments (2013, background): “Freedman (1997, 2008), Robins (1994, 2004) and Berk (2004) have drawn attention to the fact that, when such a parametric modelling assumption is false, this estimate may be extremely inaccurate and may lead to false conclusions.”
Penalized Q-Learning for Dynamic Treatment Regimens (2015, positive): “, 2000, 2002, 2007) and semiparametric methods (Murphy, 2003; Robins, 2004; Lunceford et al., 2002; Wahed and Tsiatis, 2004, 2006; Moodie et al., 2009).”
Demystifying Optimal Dynamic Treatment Regimes (2007, positive): “Under additive local rank preservation, H j ( ψ ) corresponds to a counterfactual outcome: (Robins, 2004, p. 204).”
Estimating Response-Maximized Decision Rules With Applications to Breastfeeding (2009, methodology): “…and Stephens 2007), and so a ‘default’ choice of variables to include in the model is found simply by taking the derivative of the blip function with respect to its parameters (Robins 2004): in our case, this gives at each interval ð@=@cjÞcj ¼ 1, i.e., we assume a constant expected counterfactual.”
Doubly Robust Estimation of Optimal Dynamic Treatment Regimes (2013, methodology): “If there is also assumed knowledge of the state S mechanism then a doubly robust form can be constructed (Robins [17]).”
Higher order influence functions and minimax estimation of nonlinear functionals (2008, methodology): “Robins [16] proved that b (·) is the unique function b∗ (·) minimizing Risk(b∗) ≡ Eθ [ σ−2 0 (X) {Y − [A − π0 (X)] b∗ (X)} 2 ] .”
Interactive Q-learning for Probabilities and Quantiles (2014, methodology): “…Richardson, 2010; Chakraborty et al., 2010; Song et al., 2011; Goldberg et al., 2013), Interactive Q-learning (Laber et al., 2014), g-estimation (Robins, 2004), A-learning (Murphy, 2003), regret-regression (Henderson et al., 2010), and policy search methods (Orellana et al., 2010; Zhao et al.,…”
Precision Medicine. (2019, methodology): “In the finite horizon setting, we can estimate the optimal dynamic treatment regimes through a variety of reinforcement learning techniques, including G-estimation (Robins 2004; Stephens 2016), Q- and A- learning (for an overview of these methods, see Schulte et al 2014, wherein G-learning is articu”
Statistical Inference in Dynamic Treatment Regimes (2010, methodology): “Robins (2004) notes two problems resulting from the non-differentiability of the max operator.”
TREE-BASED REINFORCEMENT LEARNING FOR ESTIMATING OPTIMAL DYNAMIC TREATMENT REGIMES (2018, background): “…[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.…”