Pith. sign in

REVIEW 3 major objections 5 minor 48 references

Doubly Robust Estimation of Causal Effect on CVR with Targeted Regularization

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper constructs a doubly robust semiparametric estimator for causal effects on post-click conversion rate that stays consistent when one of its nuisance models fails.

desk verdict The paper's advertised double robustness fails when the click model μ1 is misspecified; the empirical framework is solid but the central theory needs correction. read the letter →

arxiv 2608.13461 v1 pith:4RIAQLMY submitted 2026-08-13 cs.LG

Jiayi Dan , Bo Li , Lu Deng , Yong Wang This is my paper · ORCID
classification cs.LG MSC 62D2062G0562G20
keywords causaleffectestimationpost-clickconversionratedoublyrobustsemiparametrictheoryinfluencefunctionvonMisesexpansiontargetedregularizationcontinuoustreatment
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper's claim is that causal effects on post-click conversion rate (CVR) — the probability of conversion among users who would click under a treatment — can be estimated consistently from full observational data, not just from the clicked subset. Existing practice either trains on clicked samples, which biases any causal estimate, or debiases a prediction loss without ensuring the final estimate is unbiased. The authors derive the influence function of the estimand $\psi_a = \mathbb{E}[P(Y_2(a)=1 \mid Y_1(a)=1, X)]$, construct a one-step doubly robust estimator from its von Mises expansion, and prove it is root-n consistent and asymptotically normal when the nuisance functions (click probability, conversion probability, treatment density) converge at $o(n^{-1/4})$ or faster, so the target converges faster than any nuisance estimate. They then turn the correction into a targeted-regularization loss inside a multi-task network, and report that this estimator outperforms standard causal baselines and a naive combination of loss debiasing with causal estimators on synthetic and real data.

What carries the argument

The load-bearing object is the influence function $\phi_a$ and its von Mises expansion, which turns the estimation problem into a distributional Taylor expansion: the leading bias of the plug-in ratio $\mu_2/\mu_1$ is subtracted off by the correction term, and the leftover remainder is quadratic in nuisance errors. In the practical framework, that same correction is realized as the targeted-regularization term, which trains a low-dimensional spline parameter $\epsilon(a)$ to make the influence-function correction nearly zero, replacing the hard one-step correction with a soft regularizer and avoiding cross-fitting. The theoretical engine is the remainder bound $R_2 = O(\|\pi-\hat\pi\|(\|\mu_1-\hat\mu_1\|+\|\mu_2-\hat\mu_2\|)+\|\mu_1-\hat\mu_1\|^2+\cdots)$, which translates nuisance convergence rates into the convergence rate of the target estimate.

What would settle it

Simulate the synthetic continuous-treatment scenario from the appendix with a fixed treatment value $a$ and a known true $\psi_a$; compute the one-step estimator with cross-fitting, replicate many times, and check whether the empirical RMSE decays at an $n^{-1/2}$ rate without any kernel smoothing or discretization of $\delta(A=a)$. If the rate is slower or the variance diverges, the Dirac-based influence function is not valid for pointwise continuous-treatment inference.

Watch

Extended reading notes

Core claim

The central discovery is a closed-form influence function for the CVR causal estimand: $\phi_a(P)=\frac{\delta(A=a)}{\pi \mu_1^2}\big[(Y_2-\mu_2)\mu_1 - (Y_1-\mu_1)\mu_2\big] + \frac{\mu_2}{\mu_1} - \psi_a(P)$, where $\pi$ is the treatment density and $\mu_1, \mu_2$ are the conditional click and conversion probabilities. From the corresponding von Mises expansion, the paper obtains a one-step estimator whose error is dominated by a second-order remainder consisting of products of nuisance errors, which yields double robustness: the estimator remains consistent if one of the nuisance models is wrong, and it converges at $\sqrt{n}$ under $o(n^{-1/4})$ nuisance rates. Framing CVR this way, the paper argues, is what previous loss-debiasing CVR work lacks: unbiased loss does not imply an unbiased estimator, whereas the influence-function construction targets the estimand directly. The same remainder analysis is used to design the targeted-regularization loss for the practical multi-task framework.

Load-bearing premise

The argument hinges on treating the point-mass weight $\delta(A=a)$ in the influence function as a legitimate mathematical object for a continuous treatment; if that pointwise influence function requires smoothing to be valid, the root-n consistency claim at a fixed treatment value does not follow.

Editorial extensions

If this is right

  • Platforms can estimate CVR causal effects from observational logs over the full population, so the click-selection bias that affects standard estimators is removed.
  • Neural networks and other flexible nonparametric models can be used for the nuisance functions without sacrificing the root-n convergence of the target effect estimate.
  • The final estimate stays consistent when the conversion model or the click model is misspecified, as long as the other nuisance parts are estimated consistently.
  • The multi-task design gives decision-makers joint estimates of the treatment's effect on clicks and on post-click conversion, so a policy that boosts clicks but hurts conversion can be detected.
  • Loss-debiased CVR prediction does not by itself yield valid causal estimates; the semiparametric estimator is the component that makes the final quantity trustworthy.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the Dirac-based gradient is given rigorous footing for continuous treatments, pointwise dose-response curves for chain outcomes would be estimable at root-n without kernel smoothing, a stronger guarantee than standard kernelized continuous-treatment results; a testable middle step is to replace $\delta(A=a)$ with a kernel and compare the rates.
  • Because the remainder structure depends only on products of nuisance errors, the estimator should transfer to longer funnel chains (impression to click to purchase, or install to registration) and to ratio-type estimands generally, provided the intermediate event is observed for all units.
  • The real-data evaluation at the individual level (AUUC/QINI) sits outside the average-level theory; a natural extension is a pseudo-outcome regression for heterogeneous CVR effects, which the paper does not develop.
  • The comparison with inverse-propensity reweighting of the loss suggests that unbiasedness of a training loss is neither necessary nor sufficient for unbiasedness of an estimand; the same logic could be tested in other selection-biased prediction settings.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes a semiparametric doubly robust estimator for the causal effect on post-click conversion rate (CVR), targeting the estimand ψ_a = E[E(Y2(a)|X)/E(Y1(a)|X)] under a chain-structured outcome setting. The estimator is a one-step bias-corrected estimator built from three nuisance functions: the click probability μ1, the conversion probability μ2, and the treatment density π. The headline theoretical claims are that the estimator remains consistent even if one of the nuisance estimators is inconsistent, that it achieves root-n consistency when all nuisances converge at o_P(n^{-1/4}), and that a targeted-regularization framework inherits these properties with a slower but still fast rate. The paper also reports experiments on synthetic, semi-synthetic, and real-world data showing strong empirical performance of the proposed multi-task network.

Significance. If the theoretical claims were correct, the paper would make a useful and nontrivial contribution: a doubly robust estimator for a ratio estimand with chain-structured outcomes, with rigorous rates and a practical neural implementation. The empirical evaluation is extensive, and the targeted-regularization framework is a plausible route to stabilization. However, the central double-robustness claim is contradicted by the paper's own von Mises expansion: when only the click model μ1 is misspecified, the remainder does not vanish, so the estimator is not consistent in that case. In addition, the root-n theory for continuous treatment is not established because the influence function uses a Dirac measure without smoothing. These two issues undermine the paper's main claimed contribution.

major comments (3)
  1. [§4.2, Lemma 4.1, Theorem 4.2] The claimed double robustness property—consistency even if one of the nuisance estimators is inconsistent—is not supported by Lemma 4.1. The remainder R2 contains the term ∫ Rμ dP(x) with |Rμ| ≤ C[(μ̂1−μ1)^2 + |μ̂1−μ1||μ̂2−μ2|]. If π̂=π and μ̂2=μ2 but μ̂1 is misspecified (e.g., μ̂1≡c≠μ1), the first two integrals in R2 vanish due to the (π−π̂) factors, but the Rμ term does not vanish because it includes the unmultiplied (μ̂1−μ1)^2. In that setting the population limit of the one-step estimator is E[μ2(X,a)(2c−μ1(X,a))/c²], which differs from ψ_a=E[μ2(X,a)/μ1(X,a)] by an O(1) quantity. Thus the estimator is consistent only if μ1 is correctly specified and at least one of π or μ2 is correct; it is not consistent when only μ1 is wrong. This contradicts the abstract and §4.2 claims and propagates to Theorem 5.1 through the T2 term in its proof.
  2. [§3 and Theorem 4.2] The theoretical analysis treats the treatment A as continuous while the influence function contains the Dirac measure δ(A=a). For a continuous treatment, the one-step estimator as written is not a well-defined statistic because observations with A=a occur with probability zero, and the empirical-process term (P_n−P){φ_a(P)} would not have finite variance under a Dirac weight. The paper calls δ a 'notational device' but does not provide the smoothing, kernel, or discretization conditions under which a Dirac-based influence function supports root-n consistency. Standard continuous-treatment doubly robust estimators require kernel smoothing and attain slower rates (e.g., Kennedy et al., 2017). Theorem 4.2 therefore overstates the theoretical guarantee for exactly the continuous-treatment setting used in the experiments.
  3. [Appendix A.3, Theorem 5.1] The proof of Theorem 5.1 imports the bound ‖ε̂n−ε̌n‖=O_p(n^{-1/3}√log n) from the proof of Lemma 3 of (Nie et al., 2021) with the statement that it is independent of the target estimand, but the targeted-regularization loss R in this paper is a different functional involving μ̂1 in denominators and a cross-product of residuals. The conditions under which the spline approximation result transfers to this modified loss are not verified, so the claimed rate for the targeted-regularized estimator is not rigorously established.
minor comments (5)
  1. [§3] The definitions of μ1 and μ2 have unbalanced parentheses: the displayed formulas are missing a closing parenthesis.
  2. [Table 2] The 'Ours' row appears to contain repeated entries in the CVR and CTR columns, making the reported AMSE values ambiguous.
  3. [Theorem 4.2] The norm ∥·∥ in conditions 2–4 is not specified; the authors should state whether it is the sup-norm or the L2(P) norm, since the proof of the remainder bound uses different norms.
  4. [§6.3] The sentence 'due to the limitations of publicly available datasets, our method may be underestimated' is vague; please specify which limitation causes the underestimation and how AUUC/QINI relate to the averaged estimand defined in §3.
  5. [§4.2] The paper uses δ(A=a) in the influence function but later mentions 'kernelized IF' for continuous treatment; this connection should be made explicit when the estimator is first introduced to avoid confusion about the role of the Dirac measure.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity found; the estimator is a standard one-step/TMLE construction from the influence function, with external benchmarks and synthetic-data checks.

full rationale

The derivation chain is self-contained in the semiparametric sense. Section 4.2 defines the estimator as the plug-in term plus the empirical mean of the influence function of the target estimand, and Lemma 4.1 then derives the von Mises expansion with an explicit second-order remainder. Theorem 4.2 proves root-n consistency under o_p(n^{-1/4}) nuisance convergence and an empirical-process condition. This is the standard one-step construction: the estimator is deliberately built to cancel the first-order bias term, and the theorem verifies that the remaining second-order remainder is negligible. That is not circular reasoning; it is the normal logic of semiparametric estimation. The targeted-regularization estimator in Section 5 fits an extra parameter epsilon to the influence-function score condition, and Theorem 5.1 obtains its rate by bounding the same R2 remainder plus a spline approximation error quoted from the external prior work of Nie et al. (2021). No load-bearing assumption is justified only by the present authors' own prior work, and no fitted nuisance parameter is renamed as a prediction. The comparison in Section 6.2.3 is an empirical ablation against an IPS-loss alternative, and the CRITEO experiments are external benchmarks. The skeptical concern that Lemma 4.1's remainder contains an unmultiplied (mu_hat_1 - mu_1)^2 term, so the claimed double robustness may fail when only mu_1 is misspecified, is a correctness and robustness criticism of the paper's stated double-robustness property, not a circularity: the equations are not silently equivalent to their own inputs. No self-citation chain, definitional reduction, or ansatz-smuggling step was found.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The central derivation rests on standard semiparametric machinery plus three domain assumptions: unconfoundedness, positivity, and a bounded click rate. The targeted regularization introduces fitted hyperparameters such as alpha, beta, grid size B, and spline dimension, which are tuned on synthetic data. No new physical or causal entities are postulated.

free parameters (5)
  • beta (targeted regularization weight) = 1 (tuned values not fully reported)
    Controls strength of the influence-function correction in L_TR; Section 6.2.2 shows AMSE degrades as beta decreases.
  • alpha (propensity loss weight) = 0.5
    Balances the log-likelihood for propensity score in L_multi-task; tuned on synthetic data.
  • B (grid size for treatment density) = not specified
    Uniform discretization of [0,1] into B grids for pi(a|x); B is not reported, and the DR estimator's Dirac correction depends on how the continuous treatment is discretized.
  • K_n (spline basis count for epsilon) = K_n approximately n^{1/6}
    B-spline dimension for the targeted regularization correction; fixed by theory but not reported in experiments.
  • epsilon floor for mu1 = 1e-9
    Added to mu1 to keep the ratio positive; the small constant affects numerical stability.
assumptions (6)
  • domain assumption Unconfoundedness: A is independent of potential outcomes (Y1(a), Y2(a)) given X.
    Section 3 assumption (ii); needed for identification of psi_a as E[mu2/mu1].
  • domain assumption Overlap or positivity: pi(a|x) >= c > 0 for all x and a.
    Section 3 assumption (i); needed for the inverse propensity weights in the influence function to be bounded.
  • domain assumption Bounded click probability: mu1(x,a) >= c > 0.
    Section 3 assumption (iii); needed for the ratio and its Taylor expansion to be well-defined and for the remainder bound.
  • domain assumption Chain structure: Y2=1 implies Y1=1, and both outcomes share the same treatment A.
    Problem formulation; enables full-sample estimation of both nuisances without click selection.
  • standard math Standard semiparametric results: von Mises expansion, CLT, Slutsky, Rademacher complexity bounds, and spline approximation properties.
    Used in Lemma 4.1 and Theorems 4.2 and 5.1; these are standard but unproved background results.
  • domain assumption Nuisance estimators satisfy the required convergence rates, including o_p(n^{-1/4}) in Theorem 4.2 and the Rademacher complexity assumptions in Theorem 5.1.
    The paper cites neural network results but does not verify these rates for its own varying-coefficient architecture and CVR data.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Doubly Robust Estimation of Causal Effect on CVR with Targeted Regularization." pith.science (2026). https://pith.science/paper/4RIAQLMY

@misc{pith2026260813461,
  author       = {Pith},
  title        = {Pith review of: Doubly Robust Estimation of Causal Effect on CVR with Targeted Regularization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4RIAQLMY}},
  note         = {Machine review of arXiv:2608.13461}
}
read the original abstract

Post-click conversion rate (CVR) is a key metric in various scenarios including e-commerce and advertising, reflecting the efficiency and user experience in the second stage of the conversion process. Estimating the causal effect on CVR is therefore of great practical importance. However, directly applying existing causal inference methods to clicked samples introduces sample selection bias and increased variance due to the exclusion of non-click data. Recent studies on CVR prediction introduce "ideal loss", which optimizes model parameters using an unbiased estimate of the loss over the full sample. Nevertheless, there is no guarantee that unbiasedness of the loss implies unbiasedness of the final estimator. We revisit this challenge from the perspective of semiparametric theory. Specifically, we develop a new doubly robust causal effect estimator for chain-structured outcomes such as CVR, and derive its theoretical properties in detail. It achieves a faster convergence rate compared to nuisance parameters estimation and is therefore more robust when using flexible nonparametric estimators, including neural networks. Based on these theoretical findings, we further design a framework based on targeted regularization to improve numerical stability and practical applicability. Extensive experiments on synthetic and real-world data demonstrate the effectiveness and robustness of our method. In addition, we find that naively combining loss debiasing with standard causal estimators underperforms our method, highlighting the necessity of developing the new estimator tailored to this CVR-style objective with solid theoretical guarantees.

Figures

Figures reproduced from arXiv: 2608.13461 by the authors.

Figure 1
Figure 1. illustrates the overall model architecture. To ex￾plicitly model the dependency µ2 < µ1, we construct the prediction of µ2 as µˆ2 = ˆµ1 × µ˜2, where µ˜2 ∈ (0, 1). Mo￾tivated by the TMLE, targeted regularization should not affect the estimation of πˆ. Accordingly, gradients from the targeted regularization terms R1 and R2 are blocked from propagating to πˆ, which improves the accuracy of propensity score estimation … view at source ↗
Figure 2
Figure 2. AMSE under different settings of α. (a) Synthetic Data (b) News [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. reveals a monotonic degradation in performance as the targeted regularization weight β decreases, which further verifies the effectiveness and necessity of this component. (a) Synthetic Data (b) News [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Cumulative Gain Curve. 8. Impact Statement This paper proposes a general framework for estimating causal effects of chain-structured outcomes such as CVR. Our method can be applied to a wide range of applications, including decision-making in e-commerce and online adve…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 18 canonical work pages

  1. [1]

    International Conference on Learning Representations , year=

    Vcnet and functional targeted regularization for learning causal effects of continuous treatments , author=. International Conference on Learning Representations , year=

    work page

  2. [2]

    Advances in neural information processing systems , volume=

    Adapting neural networks for the estimation of treatment effects , author=. Advances in neural information processing systems , volume=

  3. [3]

    UCI Machine Learning Respository , volume=

    Bag of words data set , author=. UCI Machine Learning Respository , volume=

    work page

  4. [4]

    International conference on machine learning , pages=

    Learning representations for counterfactual inference , author=. International conference on machine learning , pages=. 2016 , organization=

    2016

  5. [5]

    Proceedings of the AAAI Conference on Artificial Intelligence , volume=

    Learning counterfactual representations for estimating individual dose-response curves , author=. Proceedings of the AAAI Conference on Artificial Intelligence , volume=

    work page

  6. [6]

    International conference on machine learning , pages=

    Estimating individual treatment effect: generalization bounds and algorithms , author=. International conference on machine learning , pages=. 2017 , organization=

    2017

  7. [7]

    Generalized random forests , author=

  8. [8]

    international conference on machine learning , pages=

    Recommendations as treatments: Debiasing learning and evaluation , author=. international conference on machine learning , pages=. 2016 , organization=

    2016

Show all 48 references
  1. [9]

    International Conference on Machine Learning , pages=

    Doubly robust joint learning for recommendation on data missing not at random , author=. International Conference on Machine Learning , pages=. 2019 , organization=

    2019

  2. [10]

    Companion Proceedings of the ACM Web Conference 2024 , pages=

    Entire chain uplift modeling with context-enhanced learning for intelligent marketing , author=. Companion Proceedings of the ACM Web Conference 2024 , pages=

    2024

  3. [11]

    Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages=

    Explicit feature interaction-aware uplift network for online marketing , author=. Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages=

  4. [12]

    arXiv preprint arXiv:2111.10106 , year=

    A large scale benchmark for individual treatment effect prediction and uplift modeling , author=. arXiv preprint arXiv:2111.10106 , year=

    arXiv

  5. [13]

    arXiv preprint arXiv:1802.08598 , year=

    Learning weighted representations for generalization across designs , author=. arXiv preprint arXiv:1802.08598 , year=

    arXiv

  6. [14]

    , author=

    CounterFactual Regression with Importance Sampling Weights. , author=. IJCAI , pages=. 2019 , organization=

    2019

  7. [15]

    International Conference on Artificial Intelligence and Statistics , pages=

    Counterfactual representation learning with balancing weights , author=. International Conference on Artificial Intelligence and Statistics , pages=. 2021 , organization=

    2021

  8. [16]

    Advances in Neural Information Processing Systems , volume=

    Generalization bounds for estimating causal effects of continuous treatments , author=. Advances in Neural Information Processing Systems , volume=

  9. [17]

    Proceedings of the AAAI Conference on Artificial Intelligence , volume=

    Adversarially balanced representation for continuous treatment effect estimation , author=. Proceedings of the AAAI Conference on Artificial Intelligence , volume=

  10. [18]

    2021 IEEE International Conference on Data Mining (ICDM) , pages=

    Addressing exposure bias in uplift modeling for large-scale online advertising , author=. 2021 IEEE International Conference on Data Mining (ICDM) , pages=. 2021 , organization=

    2021

  11. [19]

    Proceedings of the 28th ACM SIGKDD conference on knowledge discovery and data mining , pages=

    Descn: Deep entire space cross networks for individual treatment effect estimation , author=. Proceedings of the 28th ACM SIGKDD conference on knowledge discovery and data mining , pages=

  12. [20]

    Advances in neural information processing systems , volume=

    Causal effect inference with deep latent-variable models , author=. Advances in neural information processing systems , volume=

  13. [21]

    International conference on learning representations , year=

    GANITE: Estimation of individualized treatment effects using generative adversarial nets , author=. International conference on learning representations , year=

  14. [22]

    Advances in Neural Information Processing Systems , volume=

    Estimating the effects of continuous-valued interventions using generative adversarial networks , author=. Advances in Neural Information Processing Systems , volume=

  15. [23]

    arXiv preprint arXiv:2202.10166 , year=

    Diffusion causal models for counterfactual estimation , author=. arXiv preprint arXiv:2202.10166 , year=

    arXiv

  16. [24]

    American Economic Review , volume=

    Double/debiased/neyman machine learning of treatment effects , author=. American Economic Review , volume=. 2017 , publisher=

    2017

  17. [25]

    2018 , publisher=

    Double/debiased machine learning for treatment and structural parameters , author=. 2018 , publisher=

    2018

  18. [26]

    Biometrika , volume=

    Quasi-oracle estimation of heterogeneous treatment effects , author=. Biometrika , volume=. 2021 , publisher=

    2021

  19. [27]

    arXiv preprint arXiv:2103.04277 , year=

    Estimating heterogeneous treatment effects for general responses , author=. arXiv preprint arXiv:2103.04277 , year=

    arXiv

  20. [28]

    Handbook of statistical methods for precision medicine , pages=

    Semiparametric doubly robust targeted double machine learning: a review , author=. Handbook of statistical methods for precision medicine , pages=. 2024 , publisher=

    2024

  21. [29]

    Proceedings of the web conference 2020 , pages=

    Large-scale causal approaches to debiasing post-click conversion rate estimation with multi-task learning , author=. Proceedings of the web conference 2020 , pages=

    2020

  22. [30]

    Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval , pages=

    Enhanced doubly robust learning for debiasing post-click conversion rate estimation , author=. Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval , pages=

  23. [31]

    Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages=

    A generalized doubly robust learning framework for debiasing post-click conversion rate prediction , author=. Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages=

  24. [32]

    arXiv preprint arXiv:2205.04701 , year=

    StableDR: Stabilized doubly robust learning for recommendation on data missing not at random , author=. arXiv preprint arXiv:2205.04701 , year=

    arXiv

  25. [33]

    arXiv preprint arXiv:2203.10258 , year=

    TDR-CL: Targeted doubly robust collaborative learning for debiased recommendations , author=. arXiv preprint arXiv:2203.10258 , year=

    arXiv

  26. [34]

    Biometrika , volume=

    Semiparametric counterfactual density estimation , author=. Biometrika , volume=. 2023 , publisher=

    2023

  27. [35]

    The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval , pages=

    Entire space multi-task model: An effective approach for estimating post-click conversion rate , author=. The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval , pages=

  28. [36]

    Journal of the Royal Statistical Society Series B: Statistical Methodology , volume=

    Varying-coefficient models , author=. Journal of the Royal Statistical Society Series B: Statistical Methodology , volume=. 1993 , publisher=

    1993

  29. [37]

    The annals of Statistics , volume=

    Statistical estimation in varying coefficient models , author=. The annals of Statistics , volume=. 1999 , publisher=

    1999

  30. [38]

    Journal of the American Statistical Association , volume=

    Smoothing spline estimation for varying coefficient models with repeatedly measured dependent variables , author=. Journal of the American Statistical Association , volume=. 2001 , publisher=

    2001

  31. [39]

    2011 , publisher=

    Targeted learning: causal inference for observational and experimental data , author=. 2011 , publisher=

    2011

  32. [40]

    The annals of statistics , pages=

    Optimal global rates of convergence for nonparametric regression , author=. The annals of statistics , pages=. 1982 , publisher=

    1982

  33. [41]

    2000 , publisher=

    Asymptotic statistics , author=. 2000 , publisher=

    2000

  34. [42]

    Epidemiology , volume=

    Identifiability and exchangeability for direct and indirect effects , author=. Epidemiology , volume=. 1992 , publisher=

    1992

  35. [43]

    Biometrics , volume=

    Doubly robust estimation in missing data and causal inference models , author=. Biometrics , volume=. 2005 , publisher=

    2005

  36. [44]

    Econometrica , volume=

    Deep neural networks for estimation and inference , author=. Econometrica , volume=. 2021 , publisher=

    2021

  37. [45]

    Electronic Journal of Statistics , volume=

    Towards optimal doubly robust estimation of heterogeneous causal effects , author=. Electronic Journal of Statistics , volume=. 2023 , publisher=

    2023

  38. [46]

    2005 , publisher=

    Statistical inference for variable importance , author=. 2005 , publisher=

    2005

  39. [47]

    Proceedings of the 41st International Conference on Machine Learning , pages=

    Doubly robust causal effect estimation under networked interference via targeted learning , author=. Proceedings of the 41st International Conference on Machine Learning , pages=

  40. [48]

    Journal of the Royal Statistical Society Series B: Statistical Methodology , volume=

    Non-parametric methods for doubly robust estimation of continuous treatment effects , author=. Journal of the Royal Statistical Society Series B: Statistical Methodology , volume=. 2017 , publisher=

    2017

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.