arXiv subjects

Torsten Hothorn

Publications and source records attributed to Torsten Hothorn.

At least 19 recordsLinked to original sources

arXiv2026

Joint Estimation of Marginal and Heterogeneous Treatment Effects

Randomized clinical trials typically aim to estimate a marginal treatment effect. While covariate adjustment can improve precision, it may change the estimand in nonlinear models due to noncollapsibility, leading to conditional rather than marginal treatment effects. At the same time, identifying prognostic and predictive covariates is important for understanding treatment effect heterogeneity and informing clinical decision-making. Keeping marginal interpretability while allowing efficiency gains and assessment of heterogeneity remains a methodological challenge. In this work, we extend nonparanormal adjusted marginal inference to allow for heterogeneous treatment effects. The proposed framework embeds the marginal treatment effect directly in a joint model for the outcome and baseline covariates. This construction preserves marginal interpretability while adjusting for potentially prognostic and/or predictive covariates. The method applies to continuous, binary, ordinal, and time-to-event outcomes and allows explicit estimation and ranking of prognostic and predictive covariates on a common scale. For continuous outcomes, we show that the asymptotic variance of the marginal treatment effect measured as Cohen's $d$ is never worse and often better under covariate adjustment than without adjustment. Efficiency gains are primarily driven by prognostic effects, with realistic predictive effects contributing little additional improvement. Simulation studies confirm these findings across outcome types and demonstrate unbiased and more efficient estimation of marginal effects for Cohen's d, log-odds ratios, and log-hazard ratios. Application to an acupuncture trial demonstrates that the method reproduces the original trial findings while improving efficiency and allowing ranking of prognostic and predictive covariates.

stat.ME↗

arXiv2025

Comparing methods to assess treatment effect heterogeneity in general parametric regression models

This paper reviews and compares methods to assess treatment effect heterogeneity in the context of parametric regression models. These methods include the standard likelihood ratio tests, bootstrap likelihood ratio tests, and Goeman's global test motivated by testing whether the random effect variance is zero. We place particular emphasis on tests based on the score-residual of the treatment effect and explore different variants of tests in this class. All approaches are compared in a simulation study, and the approach based on residual scores is illustrated in a clinical trial with time-to-event outcome comparing treatment versus placebo. Our findings demonstrate that score-residual based methods provide practical, flexible and reliable tools for exploring treatment effect heterogeneity and treatment effect modifiers, and can provide useful guidance for decision making around treatment effect heterogeneity.

stat.AP↗

arXiv2025

Nonparanormal Adjusted Marginal Inference

Although treatment effects can be estimated from observed outcome distributions obtained from proper randomization in clinical trials, covariate adjustment is recommended to increase precision. For important treatment effects, such as odds or hazard ratios, conditioning on covariates in binary logistic or proportional hazards models changes the interpretation of the treatment effect and conditioning on different sets of covariates renders the resulting effect estimates incomparable. We propose a novel nonparanormal model formulation for adjusted marginal inference. This model for the joint distribution of outcome and covariates directly features a marginally defined treatment effect parameter, such as a marginal odds or hazard ratio. Not only the marginal treatment effect of interest can be estimated based on this model, it also provides an overall coefficient of determination and covariate-specific measures of prognostic strength. For the special case of Cohen's standardized mean difference d, we theoretically show that adjusting for an informative prognostic variable improves the precision of the marginal, noncollapsible effect. Empirical results confirm this not only for Cohen's d but also for odds and hazard ratios in simulations and three applications. A reference implementation is available in the R add-on package tram.

stat.ME↗

arXiv2025

Likelihood-based Modeling of Covariate-Specific Time-Dependent ROC Curves

Identifying reliable biomarkers for predicting clinical events in longitudinal studies is important for accurate disease prognosis and for guiding development of new treatments. However, prognostic studies are often observational, making it difficult to account for patient heterogeneity. In amyotrophic lateral sclerosis (ALS), factors such as age, site of onset and genetic status influence both survival and biomarker levels, yet their impact on the prognostic accuracy of biomarkers over time remains unclear. While time-dependent receiver operating characteristic methods have been developed to handle censored time-to-event outcomes, most do not adjust for covariates. To address this, we propose the nonparanormal prognostic biomarker framework, which models the joint distribution of the biomarker and event time while accounting for covariates. This allows estimation of covariate-specific time-dependent ROC curves and related summary measures. We apply the NPB framework to evaluate serum neurofilament light as a prognostic biomarker in ALS, showing that its accuracy varies over time and with patient characteristics. By capturing these covariate-specific effects, the NPB framework supports more targeted risk stratification and can potentially improve the design of clinical trials for new ALS treatments.

stat.ME↗

arXiv2024

On Nonparanormal Likelihoods

Nonparanormal models describe the joint distribution of multivariate responses via latent Gaussian, and thus parametric, copulae while allowing flexible nonparametric marginals. Some aspects of such distributions, for example conditional independence, are formulated parametrically. Other features, such as marginal distributions, can be formulated non- or semiparametrically. Such models are attractive when multivariate normality is questionable but interpretability paramount. Most estimation procedures perform two steps, first estimating the nonparametric part. The copula parameters come second, treating the marginal estimates as known. This is sufficient for some applications. For other applications, e.g. when a semiparametric margin features parameters of interest or when standard errors are important, a simultaneous estimation of all parameters might be more advantageous. We present suitable parameterisations of nonparanormal models, possibly including semiparametric effects, and define four novel nonparanormal log-likelihood functions. In general, the corresponding one-step optimisation problems are shown to be non-convex. In some cases, however, biconvex problems emerge. Several convex approximations are discussed. From a low-level computational point of view, the core contribution is the score function for multivariate normal log-probabilities computed via Genz procedure. As a demonstration for the versatility of the theoretical and computational framework, we present a series of nonparanormal models for transformation discriminant analysis when some biomarkers are subject to limit-of-detection problems. Possible empirical gains of full maximum likelihood estimation compared to two-step approaches are illustrated in a simulation study targeting semiparametric efficient polychoric correlation analysis where a theoretical benchmark is available.

stat.ME↗

arXiv2024

Smooth Transformation Models for Survival Analysis: A Tutorial Using R

Over the last five decades, we have seen strong methodological advances in survival analysis, using parametric methods and, more prominently, methods based on non-/semi-parametric estimation. As the methodological landscape continues to evolve, the task of navigating through the multitude of methods and identifying available software resources is becoming increasingly challenging -- especially in more complex scenarios, such as when dealing with interval-censored or clustered survival data, non-proportional hazards, or dependent censoring. This tutorial explores the potential of using the framework of smooth transformation models for survival analysis in the R system for statistical computing. This framework provides a unified maximum-likelihood approach that covers a wide range of survival models, including well-established ones such as the Weibull model and a fully parametric version of the famous Cox proportional hazards model, and various extensions for more complex scenarios. We explore models for non-proportional/crossing hazards, dependent censoring, clustered observations and extensions towards personalised medicine within this framework. Using survival data from a two-arm randomised controlled trial on rectal cancer therapy, we demonstrate how survival analysis tasks can be seamlessly navigated in R within this framework using the implementation provided by the "tram" package, and few related packages.

stat.ME↗

arXiv2024

Transformation Discriminant Analysis for Constructing Optimal Biomarker Combinations

Accurate diagnostic tests are essential for effective screening and treatment. However, individual biomarkers often fail to provide sufficient diagnostic accuracy, as they typically capture only one aspect of the complex disease process. Combining multiple biomarkers, each capturing a distinct mechanism, can help constructing more informative diagnostic tests. In practice, logistic regression is used as the default to combine biomarkers, but it can perform poorly when biomarker distributions exhibit skewness or differ across disease groups. Nonparametric methods provide more flexibility but generally require large sample sizes that are infrequently available in biomedical research. We propose a novel framework called transformation discriminant analysis which combines biomarkers through the likelihood ratio function to construct theoretically optimal diagnostic scores. Transformation discriminant analysis balances between flexibility and efficiency. It can accommodate a wide range of distributional shapes and disease-specific dependence structures while remaining fully parametric. This allows for likelihood inference and strong performance even in small-sample settings. We evaluate TDA through simulations and benchmark its performance against commonly used methods. Finally, we illustrate its utility in constructing an optimal diagnostic test for hepatocellular carcinoma, a disease with no single ideal biomarker. An open-source R implementation is provided for reproducibility and broader application.

stat.ME↗

arXiv2023

Model-based causal feature selection for general response types

Discovering causal relationships from observational data is a fundamental yet challenging task. Invariant causal prediction (ICP, Peters et al., 2016) is a method for causal feature selection which requires data from heterogeneous settings and exploits that causal models are invariant. ICP has been extended to general additive noise models and to nonparametric settings using conditional independence tests. However, the latter often suffer from low power (or poor type I error control) and additive noise models are not suitable for applications in which the response is not measured on a continuous scale, but reflects categories or counts. Here, we develop transformation-model (TRAM) based ICP, allowing for continuous, categorical, count-type, and uninformatively censored responses (these model classes, generally, do not allow for identifiability when there is no exogenous heterogeneity). As an invariance test, we propose TRAM-GCM based on the expected conditional covariance between environments and score residuals with uniform asymptotic level guarantees. For the special case of linear shift TRAMs, we also consider TRAM-Wald, which tests invariance based on the Wald statistic. We provide an open-source R package 'tramicp' and evaluate our approach on simulated data and in a case study investigating causal features of survival in critically ill patients.

stat.ME↗

arXiv2022

Deep conditional transformation models for survival analysis

An every increasing number of clinical trials features a time-to-event outcome and records non-tabular patient data, such as magnetic resonance imaging or text data in the form of electronic health records. Recently, several neural-network based solutions have been proposed, some of which are binary classifiers. Parametric, distribution-free approaches which make full use of survival time and censoring status have not received much attention. We present deep conditional transformation models (DCTMs) for survival outcomes as a unifying approach to parametric and semiparametric survival analysis. DCTMs allow the specification of non-linear and non-proportional hazards for both tabular and non-tabular data and extend to all types of censoring and truncation. On real and semi-synthetic data, we show that DCTMs compete with state-of-the-art DL approaches to survival analysis.

cs.LG↗

arXiv2022

Heterogeneous Treatment Effect Estimation for Observational Data using Model-based Forests

The estimation of heterogeneous treatment effects (HTEs) has attracted considerable interest in many disciplines, most prominently in medicine and economics. Contemporary research has so far primarily focused on continuous and binary responses where HTEs are traditionally estimated by a linear model, which allows the estimation of constant or heterogeneous effects even under certain model misspecifications. More complex models for survival, count, or ordinal outcomes require stricter assumptions to reliably estimate the treatment effect. Most importantly, the noncollapsibility issue necessitates the joint estimation of treatment and prognostic effects. Model-based forests allow simultaneous estimation of covariate-dependent treatment and prognostic effects, but only for randomized trials. In this paper, we propose modifications to model-based forests to address the confounding issue in observational data. In particular, we evaluate an orthogonalization strategy originally proposed by Robinson (1988, Econometrica) in the context of model-based forests targeting HTE estimation in generalized linear models and transformation models. We found that this strategy reduces confounding effects in a simulated study with various outcome distributions. We demonstrate the practical aspects of HTE estimation for survival and ordinal outcomes by an assessment of the potentially heterogeneous effect of Riluzole on the progress of Amyotrophic Lateral Sclerosis.

stat.ME↗

arXiv2022

Distribution-Free Location-Scale Regression

We introduce a generalized additive model for location, scale, and shape (GAMLSS) next of kin aiming at distribution-free and parsimonious regression modelling for arbitrary outcomes. We replace the strict parametric distribution formulating such a model by a transformation function, which in turn is estimated from data. Doing so not only makes the model distribution-free but also allows to limit the number of linear or smooth model terms to a pair of location-scale predictor functions. We derive the likelihood for continuous, discrete, and randomly censored observations, along with corresponding score functions. A plethora of existing algorithms is leveraged for model estimation, including constrained maximum-likelihood, the original GAMLSS algorithm, and transformation trees. Parameter interpretability in the resulting models is closely connected to model selection. We propose the application of a novel best subset selection procedure to achieve especially simple ways of interpretation. All techniques are motivated and illustrated by a collection of applications from different domains, including crossing and partial proportional hazards, complex count regression, non-linear ordinal regression, and growth curves. All analyses are reproducible with the help of the "tram" add-on package to the R system for statistical computing and graphics.

stat.ME↗

arXiv2022

Transformation models for ROC analysis

Receiver operating characteristic (ROC) analysis is one of the most popular approaches for evaluating and comparing the accuracy of medical diagnostic tests. Although various methodologies have been developed for estimating ROC curves and its associated summary indices, there is no consensus on a single framework that can provide consistent statistical inference whilst handling the complexities associated with medical data. Such complexities might include covariates that influence the diagnostic potential of a test, ordinal test data, censored data due to instrument detection limits or correlated biomarkers. We propose a regression model for the transformed test results which exploits the invariance of ROC curves to monotonic transformations and naturally accommodates these features. Our use of maximum likelihood inference guarantees asymptotic efficiency of the resulting estimators and associated confidence intervals. Simulation studies show that the estimates based on transformation models are unbiased and yield coverage at nominal levels. The methodology is applied to a cross-sectional study of metabolic syndrome where we investigate the covariate-specific performance of weight-to-height ratio as a non-invasive diagnostic test. Software implementations for all the methods described in the article are provided in the "tram" R package.

stat.ME↗

arXiv2022

What Makes Forest-Based Heterogeneous Treatment Effect Estimators Work?

Estimation of heterogeneous treatment effects (HTE) is of prime importance in many disciplines, ranging from personalized medicine to economics among many others. Random forests have been shown to be a flexible and powerful approach to HTE estimation in both randomized trials and observational studies. In particular "causal forests", introduced by Athey, Tibshirani and Wager (2019), along with the R implementation in package grf were rapidly adopted. A related approach, called "model-based forests", that is geared towards randomized trials and simultaneously captures effects of both prognostic and predictive variables, was introduced by Seibold, Zeileis and Hothorn (2018) along with a modular implementation in the R package model4you. Here, we present a unifying view that goes beyond the theoretical motivations and investigates which computational elements make causal forests so successful and how these can be blended with the strengths of model-based forests. To do so, we show that both methods can be understood in terms of the same parameters and model assumptions for an additive model under L2 loss. This theoretical insight allows us to implement several flavors of "model-based causal forests" and dissect their different elements in silico. The original causal forests and model-based forests are compared with the new blended versions in a benchmark study exploring both randomized trials and observational settings. In the randomized setting, both approaches performed akin. If confounding was present in the data generating process, we found local centering of the treatment indicator with the corresponding propensities to be the main driver for good performance. Local centering of the outcome was less important, and might be replaced or enhanced by simultaneous split selection with respect to both prognostic and predictive effects.

stat.ME↗

arXiv2022

Deep interpretable ensembles

Ensembles improve prediction performance and allow uncertainty quantification by aggregating predictions from multiple models. In deep ensembling, the individual models are usually black box neural networks, or recently, partially interpretable semi-structured deep transformation models. However, interpretability of the ensemble members is generally lost upon aggregation. This is a crucial drawback of deep ensembles in high-stake decision fields, in which interpretable models are desired. We propose a novel transformation ensemble which aggregates probabilistic predictions with the guarantee to preserve interpretability and yield uniformly better predictions than the ensemble members on average. Transformation ensembles are tailored towards interpretable deep transformation models but are applicable to a wider range of probabilistic neural networks. In experiments on several publicly available data sets, we demonstrate that transformation ensembles perform on par with classical deep ensembles in terms of prediction performance, discrimination, and calibration. In addition, we demonstrate how transformation ensembles quantify both aleatoric and epistemic uncertainty, and produce minimax optimal predictions under certain conditions.

stat.ML↗

arXiv2022

Joint Count Transformation Models with Covariate-dependent Correlations

Joint Species Distribution Models are essential for understanding how ecological covariates shape species communities. However, most existing approaches are limited by rigid parametric distributions for count data and the inability to model how interspecific associations change with those covariates. We introduce joint count transformation models, a novel framework designed to overcome these limitations. Our approach combines distribution-free marginal count transformation models for multiple species with a covariate-dependent latent Gaussian copula to model interspecific correlations, interpretable as Spearman's rank correlation on the observed count scale. All model parameters are estimated efficiently via joint maximum likelihood estimation, implemented in the R package tram. We apply this framework to model the joint abundance of three fish-eating bird species, using seasonality as the primary covariate. Our model successfully captured the complex, species-specific seasonal abundance patterns, including periods of high zero-counts and seasonal shifts in variance. Furthermore, the model revealed strong, seasonally-varying correlations between the species. These findings are consistent with an empirical approach and similar to those from the computationally expensive parametric Bayesian Hierarchical Modelling of Species Communities (HMSC) framework. Consistency, accuracy and feasibility of our approach are demonstrated in a simulation study for up to 10 species.

stat.ME↗

arXiv2021

On the relevance of prognostic information for clinical trials: A theoretical quantification

The question of how individual patient data from cohort studies or historical clinical trials can be leveraged for designing more powerful, or smaller yet equally powerful, clinical trials becomes increasingly important in the era of digitalisation. Today, the traditional statistical analyses approaches may seem questionable to practitioners in light of ubiquitous historical covariate information. Several methodological developments aim at incorporating historical information in the design and analysis of future clinical trials, most importantly Bayesian information borrowing, propensity score methods, stratification, and covariate adjustment. Recently, adjusting the analysis with respect to a prognostic score, which was obtained from some machine learning procedure applied to historical data, has been suggested and we study the potential of this approach for randomised clinical trials. In an idealised situation of a normal outcome in a two-arm trial with 1:1 allocation, we derive a simple sample size reduction formula as a function of two criteria characterising the prognostic score: (1) The coefficient of determination $R^2$ on historical data and (2) the correlation $\rho$ between the estimated and the true unknown prognostic scores. While maintaining the same power, the original total sample size $n$ planned for the unadjusted analysis reduces to $(1 - R^2 \rho^2) \times n$ in an adjusted analysis. Robustness in less ideal situations was assessed empirically. We conclude that there is potential for substantially more powerful or smaller trials, but only when prognostic scores can be accurately estimated.

stat.ME↗

arXiv2021

Probabilistic Time Series Forecasts with Autoregressive Transformation Models

Probabilistic forecasting of time series is an important matter in many applications and research fields. In order to draw conclusions from a probabilistic forecast, we must ensure that the model class used to approximate the true forecasting distribution is expressive enough. Yet, characteristics of the model itself, such as its uncertainty or its feature-outcome relationship are not of lesser importance. This paper proposes Autoregressive Transformation Models (ATMs), a model class inspired by various research directions to unite expressive distributional forecasts using a semi-parametric distribution assumption with an interpretable model specification. We demonstrate the properties of ATMs both theoretically and through empirical evaluation on several simulated and real-world forecasting datasets.

cs.LG↗

arXiv2020

Deep and interpretable regression models for ordinal outcomes

Outcomes with a natural order commonly occur in prediction tasks and often the available input data are a mixture of complex data like images and tabular predictors. Deep Learning (DL) models are state-of-the-art for image classification tasks but frequently treat ordinal outcomes as unordered and lack interpretability. In contrast, classical ordinal regression models consider the outcome's order and yield interpretable predictor effects but are limited to tabular data. We present ordinal neural network transformation models (ONTRAMs), which unite DL with classical ordinal regression approaches. ONTRAMs are a special case of transformation models and trade off flexibility and interpretability by additively decomposing the transformation function into terms for image and tabular data using jointly trained neural networks. The performance of the most flexible ONTRAM is by definition equivalent to a standard multi-class DL model trained with cross-entropy while being faster in training when facing ordinal outcomes. Lastly, we discuss how to interpret model components for both tabular and image data on two publicly available datasets.

stat.ML↗