Tessera

Lars van der Laan

@larsvanderlaan3.bsky.social

Postdoc @Stanford | Ph.D. @UW Statistics | ML Research @Netflix — Machine learning, semiparametric statistics, causal inference, and reinforcement learning. https://larsvanderlaan.github.io

30 Posts598 Follower 137 Folgt

ArXiv Paperboy (Stat.ME+Econ.EM)@paperposterbot.bsky.social · last wk.

arXiv📈🤖 The V-fold jackknife for semiparametric inference: variance estimation, confidence intervals, and simultaneous confidence bands By Li, Ertefaie, Laan

$For decades, the bootstrap has been a default tool for statistical inference because of its broad applicability and minimal analytic requirements. Although its validity is well understood for smooth parametric estimators, its theoretical properties for many modern semiparametric and machine-learning estimators remain largely unstudied. Nevertheless, bootstrap procedures are often used routinely in such settings, even when their validity is unknown and their computational cost is substantial. We develop the $V$-fold jackknife as a computationally efficient and theoretically justified alternative for semiparametric inference. It requires only $V$ leave-fold-out refits and uses the empirical dispersion of jackknife pseudo-values to quantify uncertainty, without deriving or evaluating an influence function. For regular asymptotically linear estimators of pathwise differentiable parameters, we show that, for fixed $V$, the Studentized $V$-fold jackknife statistic converges to a $t$-distribution with $V-1$ degrees of freedom, giving valid confidence intervals even though the jackknife variance estimator does not converge in probability. When $V\to\infty$, we establish consistency of the variance estimator at rate $V^{-1/2}$, allowing $V$ to diverge slowly, for example at rate $\log n$. We also develop simultaneous confidence bands based on the correct componentwise-Studentized limiting distribution. Finally, we extend the theory to generalized asymptotically linear estimators with diverging influence-function variance and slower-than-$\sqrt n$ convergence; scale invariance of Studentization eliminates the need to know the effective convergence rate. Simulations on the average treatment effect, Kaplan--Meier survival curve, and highly adaptive lasso dose-response curves confirm reliable inference, including where influence-function-based standard errors are anti-conservative or unstable.$

arXiv stat.ML Machine Learning @statml-bot.bsky.social · 4w ago

Lars van der Laan, Nathan Kallus: Fitted Occupancy-Ratio Evaluation without Bellman Completeness https://arxiv.org/abs/2607.05375 https://arxiv.org/pdf/2607.05375 https://arxiv.org/html/2607.05375

ArXiv Paperboy (Stat.ME+Econ.EM)@paperposterbot.bsky.social · 2mo ago

arXiv📈🤖 AI-Assisted Variance Reduction in Randomized Experiments By Arbour, Ben-Michael, Feller et al

Generative AI and large language models can produce realistic predictions of human behavior from rich, unstructured inputs with little to no task-specific training data. Recent work uses these ``digital twin'' predictions to supplement human responses in surveys and experiments. We study the special case of using AI-generated predictions to reduce variance in randomized experiments. We argue that doing so requires no new estimators and that researchers can simply include AI predictions as covariates in standard regression adjustment, analogous to adjusting for a prognostic score. A benefit of this approach is a ``do no harm'' property whereby the adjusted estimator reverts to the unadjusted difference in means when predictions are uninformative. Other methods, such as variants of prediction-powered inference, do not have this guarantee. We provide implementation guidance, including how to obtain continuous scores from discrete LLM outputs and how to use LLMs to featurize unstructured inputs as auxiliary covariates. We demonstrate these ideas in simulations and three empirical applications: a survey mega-study, an email marketing A/B test, and a large-scale technology platform experiment. Overall, efficiency gains are real if modest, with greater benefits in studies that contain substantial text and other unstructured data. We also confirm the do no harm property empirically. Given these gains and limited costs, we recommend adjusting for AI-generated predictions as a regular empirical practice.

arxiv stat.ML @arxiv-stat-ml.bsky.social · 2mo ago

Nicolas Emmenegger, Ellery Stahler, Chara Podimata Prediction-Powered Inference Across Many Tasks for AI Evaluation & Social Science Research https://arxiv.org/abs/2605.29249

Michael Schomaker @mfschomaker.bsky.social · 2mo ago

1/ I have never really advertised our little Github-R-package "SLbooster": Additional and Modified Learning and Screening functions for Super Learning: github.com/MichaelSchom... What does it do? 👇

GitHub - MichaelSchomaker/SLbooster

Contribute to MichaelSchomaker/SLbooster development by creating an account on GitHub.

github.com

Stefan Feuerriegel @sfeuerriegel.bsky.social · 2mo ago

📢Checkout our new overview on #CausalML in Wiley #StatsRef 👉We give a concise overview of ML for causal inference — incl. IPTW, AIPTW, TMLE, meta-learners, Neyman orthogonality, ... 📄 doi.org/10.1002/9781... (or PM me) with @larsvanderlaan3.bsky.social @valik-melnychuk.bsky.social

ArXiv Paperboy (Stat.ME+Econ.EM)@paperposterbot.bsky.social · 3mo ago

arXiv📈🤖 Calibeating Prediction-Powered Inference By Laan, Laan

We study semisupervised mean estimation with a small labeled sample, a large unlabeled sample, and a black-box prediction model whose output may be miscalibrated. A standard approach in this setting is augmented inverse-probability weighting (AIPW) [Robins et al., 1994], which protects against prediction-model misspecification but can be inefficient when the prediction score is poorly aligned with the outcome scale. We introduce Calibrated Prediction-Powered Inference, which post-hoc calibrates the prediction score on the labeled sample before using it for semisupervised estimation. This simple step requires no retraining and can improve the original score both as a predictor of the outcome and as a regression adjustment for semisupervised inference. We study both linear and isotonic calibration. For isotonic calibration, we establish first-order optimality guarantees: isotonic post-processing can improve predictive accuracy and estimator efficiency relative to the original score and simpler post-processing rules, while no further post-processing of the fitted isotonic score yields additional first-order gains. For linear calibration, we show first-order equivalence to PPI++. We also clarify the relationship among existing estimators, showing that the original PPI estimator is a special case of AIPW and can be inefficient when the prediction model is accurate, while PPI++ is AIPW with empirical efficiency maximization [Rubin et al., 2008]. In simulations and real-data experiments, our calibrated estimators often outperform PPI and are competitive with, or outperform, AIPW and PPI++. We provide an accompanying Python package, ppi_aipw, at https://larsvanderlaan.github.io/ppi-aipw/.

Lars van der Laan @larsvanderlaan3.bsky.social · 5mo ago

A very nice overview of PPI, which is closely related to the AIPW estimator from missing data and the even older model-assisted estimators from survey sampling.

ArXiv Paperboy (Stat.ME+Econ.EM)@paperposterbot.bsky.social · 5mo ago

arXiv📈🤖 PPI is the Difference Estimator: Recognizing the Survey Sampling Roots of Prediction-Powered Inference By Mozer

Prediction-powered inference (PPI) is a rapidly growing framework for combining machine learning predictions with a small set of gold-standard labels to conduct valid statistical inference. In this article, I argue that the core estimators underlying PPI are equivalent to well-established estimators from the survey sampling literature dating back to the 1970s. Specifically, the PPI estimator for a population mean is algebraically equivalent to the difference estimator of Cassel et al. (1976), and PPI plus corresponds to the generalized regression (GREG) estimator of Sarndal et al. (2003). Recognizing this equivalence, I consider what part of PPI is inherited from a long-standing literature in statistics, what part is genuinely new, and where inferential claims require care. After introducing the two frameworks and establishing their equivalence, I break down where PPI diverges from model-assisted estimation, including differences in the mode of inference, the role of the unlabeled data pool, and the consequences of differential prediction error for subgroup estimands such as the average treatment effect. I then identify what each framework offers the other: PPI researchers can draw on the survey sampling literature's well-developed theory of calibration, optimal allocation, and design-based diagnostics, while survey sampling researchers can benefit from PPI's extensions to non-standard estimands and its accessible software ecosystem. The article closes with a call for integration between these two communities, motivated by the growing use of large language models as measurement instruments in applied research.

Lars van der Laan @larsvanderlaan3.bsky.social · 5mo ago

Quite happy with this updated version of an older preprint of mine. The paper gives a unifying perspective on many recent adaptive estimators in causal inference, connecting DML, post-model-selection inference, and superefficiency. Some neat connections to calibrated DML and automatic DML.

ArXiv Paperboy (Stat.ME+Econ.EM)@paperposterbot.bsky.social · 5mo ago

arXiv📈🤖 Adaptive debiased machine learning using data-driven model selection techniques By Laan, Carone, Luedtke et al

$Debiased machine learning estimators for smooth functionals in nonparametric models can exhibit substantial variability and instability, often leading practitioners to instead rely on parametric or semiparametric working models. Such models, however, may be misspecified and can therefore introduce bias. We study how data-driven model selection can be combined with debiased machine learning to construct estimators that adapt to structure in the data-generating distribution. To this end, we propose Adaptive Debiased Machine Learning (ADML), a nonparametric framework for constructing superefficient estimators of pathwise differentiable parameters. The framework unifies a broad class of previously proposed adaptive estimators, including methods based on variable selection, learned feature representations, and collaborative targeted learning. It requires only high-level conditions and approximate validity of the selection procedure, which are implied by lower-level conditions already assumed in important settings, including sieve-based selection, sparsity-based methods such as the Lasso, and data-adaptive feature representations. We show that ADML estimators yield regular and efficient root-$n$ inference for an oracle projection parameter induced by a data-adaptive oracle submodel. This oracle parameter coincides with the target parameter at the true distribution but typically has a smaller efficiency bound, thereby yielding superefficiency for the target parameter. As a practical illustration, we introduce a broad class of automatic ADML estimators for continuous linear functionals of the outcome regression, in which model selection is performed directly on the regression itself. Motivated by overlap challenges in causal inference, we develop new superefficient plug-in estimators for the average treatment effect based on calibration in semiparametric regression models.$

Lars van der Laan @larsvanderlaan3.bsky.social · 5mo ago

🚨A Researcher's Guide to Empirical Risk Minimization I put together a guide on regret theory for empirical risk minimization (ERM) as I understand it. The goal was to compile results and proof techniques I’ve found useful in my own work. I hope people find it useful more broadly

arXiv stat.ML Machine Learning@statml-bot.bsky.social · 5mo ago

Lars van der Laan: A Researcher's Guide to Empirical Risk Minimization https://arxiv.org/abs/2602.21501 https://arxiv.org/pdf/2602.21501 https://arxiv.org/html/2602.21501

Lars van der Laan @larsvanderlaan3.bsky.social · 7mo ago

New Paper: Efficient Inference for IRL & Dynamic Discrete Choice We study reward recovery from behavior and inference in inverse RL and DDC, w/o parametric restrictions, while also simplifying optimization A semiparametric extension of the influential Rust (1987) paper arxiv.org/pdf/2512.24407

arxiv.org

arxiv stat.ML @arxiv-stat-ml.bsky.social · 7mo ago

Lars van der Laan, Nathan Kallus Stationary Reweighting Yields Local Convergence of Soft Fitted Q-Iteration https://arxiv.org/abs/2512.23927

arxiv stat.ML @arxiv-stat-ml.bsky.social · 7mo ago

Lars van der Laan, Nathan Kallus Fitted Q Evaluation Without Bellman Completeness via Stationary Weighting https://arxiv.org/abs/2512.23805

ArXiv Paperboy (Stat.ME+Econ.EM)@paperposterbot.bsky.social · 7mo ago

link 📈🤖 Bellman Calibration for V-Learning in Offline Reinforcement Learning (Laan, Kallus) We introduce Iterated Bellman Calibration, a simple, model-agnostic, post-hoc procedure for calibrating off-policy value predictions in infinite-horizon Markov decision processes. Bellman calibration requi

Aleksander Molak @alxndrmlk.bsky.social · 11mo ago

He did it before Double Machine Learning I met with professor Mark van der Laan because I think his work is pretty incredible and it sometimes feels like a secret that only a few people know about, especially in industry. 1/ #CausalSky #StatSky #CausalInference

Alex Luedtke @alexluedtke.bsky.social · last yr.

I've advised 15 PhD students—10 were international students. All graduates continue advancing U.S. excellence in research and education. Cutting off this pipeline of talent would be shortsighted.

apoorva lal @apoorvalal.com · last yr.

I had a hard time believing it was as simple as this until Lars taught me how to implement it - calibrate=True and you're done github.com/apoorvalal/a...

GitHub - apoorvalal/aipyw: minimal, fast, object-oriented implementation of the AIPW and related estimators for many discrete treatments. Implemented with scikitlearners and cross-fitting.

minimal, fast, object-oriented implementation of the AIPW and related estimators for many discrete treatments. Implemented with scikitlearners and cross-fitting. - GitHub - apoorvalal/aipyw: minim...

github.com

Lars van der Laan@larsvanderlaan3.bsky.social · last yr.

Had a great time presenting at #ACIC on doubly robust inference via calibration Calibrating nuisance estimates in DML protects against model misspecification and slow convergence. Just one line of code is all it takes.

Lars van der Laan @larsvanderlaan3.bsky.social · last yr.

ArXiv Paperboy (Stat.ME+Econ.EM)@paperposterbot.bsky.social · last yr.

link 📈🤖 Nonparametric Instrumental Variable Inference with Many Weak Instruments (Laan, Kallus, Bibaut) We study inference on linear functionals in the nonparametric instrumental variable (NPIV) problem with a discretely-valued instrument under a many-weak-instruments asymptotic regime, where the

Lars van der Laan @larsvanderlaan3.bsky.social · last yr.

I’ll be giving an oral presentation at ACIC in the Advancing Causal Inference session with ML on Wednesday! My talk will be on Automatic Double Reinforcement Learning and long term causal inference! I’ll discuss Markov decision processes, Q-functions, and a new form of calibration for RL!

Lars van der Laan @larsvanderlaan3.bsky.social · last yr.

New preprint with #Netflix out! We study the NPIV problem with a discrete instrument under a many-weak-instruments regime. A key application: constructing confounding-robust surrogates using past experiments as instruments. My mentor Aurélien Bibaut will be presenting a poster at #ACIC2025!

arxiv.stat.ME@arxiv-stat-me.bsky.social · last yr.

Lars van der Laan, Nathan Kallus, Aur\'elien Bibaut Nonparametric Instrumental Variable Inference with Many Weak Instruments https://arxiv.org/abs/2505.07729

Lars van der Laan @larsvanderlaan3.bsky.social · last yr.

Our work on stabilized inverse probability weighting via calibration was accepted to #CLeaR2025! I gave an oral presentation last week and was honored to receive the Best Paper Award. I’ll be giving a related poster talk at #ACIC on calibration and DML and how it provides doubly robust inference!

arxiv.stat.ME@arxiv-stat-me.bsky.social · 2y ago

Lars van der Laan, Ziming Lin, Marco Carone, Alex Luedtke Stabilized Inverse Probability Weighting via Isotonic Calibration https://arxiv.org/abs/2411.06342

Lars van der Laan @larsvanderlaan3.bsky.social · last yr.

Valeriy M., PhD, MBA, CQF @predict-addict.bsky.social · last yr.

The paper "Generalized Venn and Venn-Abers Calibration with Applications in Conformal Prediction" by Lars van der Laan and Ahmed Alaa introduces a comprehensive framework that extends Venn and Venn-Abers calibration methods to a broad range of prediction tasks and loss functions.

Lars van der Laan @larsvanderlaan3.bsky.social · last yr.

🚨 Excited about this new paper on Generalized Venn Calibration and conformal prediction! We show that Venn and Venn-Abers can be extended to general losses, and that conformal prediction can be viewed as Venn multicalibration for the quantile loss! #calibration #conformal

arxiv stat.ML@arxiv-stat-ml.bsky.social · last yr.

Lars van der Laan, Ahmed Alaa Generalized Venn and Venn-Abers Calibration with Applications in Conformal Prediction https://arxiv.org/abs/2502.05676

arxiv stat.ML @arxiv-stat-ml.bsky.social · last yr.

Lars van der Laan, Ahmed Alaa Generalized Venn and Venn-Abers Calibration with Applications in Conformal Prediction https://arxiv.org/abs/2502.05676

Iván Díaz @idiaz.bsky.social · 2y ago

Your comment also reminds me of this paper where they ensure the estimators solve a certain equation (which I think can be viewed as a kind of balance) using isotonic regression and they show this leads to DR inference: arxiv.org/pdf/2411.02771

arxiv.org

Lars van der Laan @larsvanderlaan3.bsky.social · 2y ago

Thrilled to share our new paper! We introduce a generalized autoDML framework for smooth functionals in general M-estimation problems, significantly broadening the scope of problems where automatic debiasing can be applied!

arxiv.stat.ME@arxiv-stat-me.bsky.social · 2y ago

Lars van der Laan, Aurelien Bibaut, Nathan Kallus, Alex Luedtke Automatic Debiased Machine Learning for Smooth Functionals of Nonparametric M-Estimands https://arxiv.org/abs/2501.11868

ArXiv Paperboy (Stat.ME+Econ.EM)@paperposterbot.bsky.social · 2y ago

link 📈🤖 Automatic Debiased Machine Learning for Smooth Functionals of Nonparametric M-Estimands (Laan, Bibaut, Kallus et al) We propose a unified framework for automatic debiased machine learning (autoDML) to perform inference on smooth functionals of infinite-dimensional M-estimands, defined as

Lars van der Laan @larsvanderlaan3.bsky.social · 2y ago

arxiv.stat.ME@arxiv-stat-me.bsky.social · 2y ago

Lars van der Laan, Aurelien Bibaut, Nathan Kallus, Alex Luedtke Automatic Debiased Machine Learning for Smooth Functionals of Nonparametric M-Estimands https://arxiv.org/abs/2501.11868