arXiv subjects

Soojin Park

Publications and source records attributed to Soojin Park.

9 recordsLinked to original sources

arXiv2026

Causal Decomposition Analysis with Synergistic Interventions: A Triply-Robust Machine Learning Approach to Addressing Multiple Dimensions of Social Disparities

Educational disparities are rooted in and perpetuate social inequalities across multiple dimensions such as race, socioeconomic status, and geography. To reduce disparities, most intervention strategies focus on a single domain and frequently evaluate their effectiveness by using causal decomposition analysis. However, a growing body of research suggests that single-domain interventions may be insufficient for individuals marginalized on multiple fronts. While interventions across multiple domains are increasingly proposed, there is limited guidance on appropriate methods for evaluating their effectiveness. To address this gap, we develop an extended causal decomposition analysis that simultaneously targets multiple causally ordered intervening factors, allowing for the assessment of their synergistic effects. These scenarios often involve challenges related to model misspecification due to complex interactions among group categories, intervening factors, and their confounders with the outcome. To mitigate these challenges, we introduce a triply robust estimator that leverages machine learning techniques to address potential model misspecification. We apply our method to a cohort of students from the High School Longitudinal Study, focusing on math achievement disparities between Black, Hispanic, and White high schoolers. Specifically, we examine how two sequential interventions - equalizing the proportion of students who attend high-performing schools and equalizing enrollment in Algebra I by 9th grade across racial groups - may reduce these disparities.

stat.ME↗

arXiv2025

Let the Experts Speak: Improving Survival Prediction & Calibration via Mixture-of-Experts Heads

Deep mixture-of-experts models have attracted a lot of attention for survival analysis problems, particularly for their ability to cluster similar patients together. In practice, grouping often comes at the expense of key metrics such as calibration error and predictive accuracy. This is due to the restrictive inductive bias that mixture-of-experts imposes, that predictions for individual patients must look like predictions for the group they're assigned to. Might we be able to discover patient group structure, where it exists, while improving calibration and predictive accuracy? In this work, we introduce several discrete-time deep mixture-of-experts (MoE)-based architectures for survival analysis problems, one of which achieves all desiderata: clustering, calibration, and predictive accuracy. We show that a key differentiator between this array of MoEs is how expressive their experts are. We find that more expressive experts that tailor predictions per patient outperform experts that rely on fixed group prototypes.

cs.LG↗

arXiv2025

Simulation-Based Sensitivity Analysis in Optimal Treatment Regimes and Causal Decomposition with Individualized Interventions

Causal decomposition analysis aims to assess the effect of modifying risk factors on reducing social disparities in outcomes. Recently, this analysis has incorporated individual characteristics when modifying risk factors by utilizing optimal treatment regimes (OTRs). Since the newly defined individualized effects rely on the no omitted confounding assumption, developing sensitivity analyses to account for potential omitted confounding is essential. Moreover, OTRs and individualized effects are primarily based on binary risk factors, and no formal approach currently exists to benchmark the strength of omitted confounding using observed covariates for binary risk factors. To address this gap, we extend a simulation-based sensitivity analysis that simulates unmeasured confounders, addressing two sources of bias emerging from deriving OTRs and estimating individualized effects. Additionally, we propose a formal bounding strategy that benchmarks the strength of omitted confounding for binary risk factors. Using the High School Longitudinal Study 2009 (HSLS:09), we demonstrate this sensitivity analysis and benchmarking method.

stat.ML↗

arXiv2025

When Do Traditional and Causal Decomposition Methods Diverge? A Practical Guide for Studying Health Disparities

Identifying malleable factors that can reduce social disparities is a central objective among researchers across disciplines. Traditionally, researchers have relied on the difference-in-coefficients and Kitagawa-Oaxaca-Blinder frameworks. More recently, methods grounded in the potential outcomes framework have emerged. While these methods share the same goal of identifying drivers of disparity, they frequently yield divergent results depending on the underlying confounding structures and research settings. Despite these significant differences, applied researchers lack clear guidance on selecting appropriate methods for their specific research contexts. To address this gap, this study provides a systematic review and offer an intuitive guidance through Directed Acyclic Graphs and comparative simulation studies. We begin by reviewing each method assuming no unmeasured confounding, which is often violated in observational settings. Consequently, we extend our analysis to two realistic scenarios: 1) unmeasured confounding exists in the relationship between intermediate confounders and the mediator, and 2) unmeasured confounding exists in the relationship between the mediator and the outcome. Finally, we illustrate these recommendations through a case study examining the role of educational attainment in explaining racial disparities in later-life cognition.

stat.AP↗

arXiv2024

Evaluation of General Large Language Models in Contextually Assessing Semantic Concepts Extracted from Adult Critical Care Electronic Health Record Notes

The field of healthcare has increasingly turned its focus towards Large Language Models (LLMs) due to their remarkable performance. However, their performance in actual clinical applications has been underexplored. Traditional evaluations based on question-answering tasks don't fully capture the nuanced contexts. This gap highlights the need for more in-depth and practical assessments of LLMs in real-world healthcare settings. Objective: We sought to evaluate the performance of LLMs in the complex clinical context of adult critical care medicine using systematic and comprehensible analytic methods, including clinician annotation and adjudication. Methods: We investigated the performance of three general LLMs in understanding and processing real-world clinical notes. Concepts from 150 clinical notes were identified by MetaMap and then labeled by 9 clinicians. Each LLM's proficiency was evaluated by identifying the temporality and negation of these concepts using different prompts for an in-depth analysis. Results: GPT-4 showed overall superior performance compared to other LLMs. In contrast, both GPT-3.5 and text-davinci-003 exhibit enhanced performance when the appropriate prompting strategies are employed. The GPT family models have demonstrated considerable efficiency, evidenced by their cost-effectiveness and time-saving capabilities. Conclusion: A comprehensive qualitative performance evaluation framework for LLMs is developed and operationalized. This framework goes beyond singular performance aspects. With expert annotations, this methodology not only validates LLMs' capabilities in processing complex medical data but also establishes a benchmark for future LLM evaluations across specialized domains.

cs.CL↗

arXiv2022

Estimation and Sensitivity Analysis for Causal Decomposition in Heath Disparity Research

In the field of disparities research, there has been growing interest in developing a counterfactual-based decomposition analysis to identify underlying mediating mechanisms that help reduce disparities in populations. Despite rapid development in the area, most prior studies have been limited to regression-based methods, undermining the possibility of addressing complex models with multiple mediators and/or heterogeneous effects. We propose an estimation method that effectively addresses complex models. Moreover, we develop a novel sensitivity analysis for possible violations of identification assumptions. The proposed method and sensitivity analysis are demonstrated with data from the Midlife Development in the US study to investigate the degree to which disparities in cardiovascular health at the intersection of race and gender would be reduced if the distributions of education and perceived discrimination were the same across intersectional groups.

stat.ME↗

arXiv2022

Sensitivity Analysis for Causal Decomposition Analysis: Assessing Robustness Toward Omitted Variable Bias

A key objective of decomposition analysis is to identify a factor (the 'mediator') contributing to disparities in an outcome between social groups. In decomposition analysis, a scholarly interest often centers on estimating how much the disparity (e.g., health disparities between Black women and White men) would be reduced/remain if we set the mediator (e.g., education) distribution of one social group equal to another. However, causally identifying disparity reduction and remaining depends on the no omitted mediator-outcome confounding assumption, which is not empirically testable. Therefore, we propose a set of sensitivity analyses to assess the robustness of disparity reduction to possible unobserved confounding. We provide sensitivity analysis techniques based on regression coefficients and $R^2$ values. The proposed techniques are flexible to address unobserved confounding measured before and after the group status. In addition, $R^2$-based sensitivity analysis offers a straightforward interpretation of sensitivity parameters and a standard way to report the robustness of research findings. Although we introduce sensitivity analysis techniques in the context of decomposition analysis, they can be utilized in any mediation setting based on interventional indirect effects.

stat.ME↗

arXiv2021

Choosing an Optimal Method for Causal Decomposition Analysis: A Better Practice for Identifying Contributing Factors to Health Disparities

Causal decomposition analysis provides a way to identify mediators that contribute to health disparities between marginalized and non-marginalized groups. In particular, the degree to which a disparity would be reduced or remain after intervening on a mediator is of interest. Yet, estimating disparity reduction and remaining might be challenging for many researchers, possibly because there is a lack of understanding of how each estimation method differs from other methods. In addition, there is no appropriate estimation method available for a certain setting (i.e., a regression-based approach with a categorical mediator). Therefore, we review the merits and limitations of the existing three estimation methods (i.e., regression, weighting, and imputation) and provide two new extensions that are useful in practical settings. A flexible new method uses an extended imputation approach to address a categorical and continuous mediator or outcome while incorporating any nonlinear relationships. A new regression method provides a simple estimator that performs well in terms of bias and variance but at the cost of assuming linearity, except for exposure and mediator interactions. Recommendations are given for choosing methods based on a review of different methods and simulation studies. We demonstrate the practice of choosing an optimal method by identifying mediators that reduce race and gender disparity in cardiovascular health, using data from the Midlife Development in the US study.

stat.ME↗

arXiv2019

Robustness of Object Recognition under Extreme Occlusion in Humans and Computational Models

Most objects in the visual world are partially occluded, but humans can recognize them without difficulty. However, it remains unknown whether object recognition models like convolutional neural networks (CNNs) can handle real-world occlusion. It is also a question whether efforts to make these models robust to constant mask occlusion are effective for real-world occlusion. We test both humans and the above-mentioned computational models in a challenging task of object recognition under extreme occlusion, where target objects are heavily occluded by irrelevant real objects in real backgrounds. Our results show that human vision is very robust to extreme occlusion while CNNs are not, even with modifications to handle constant mask occlusion. This implies that the ability to handle constant mask occlusion does not entail robustness to real-world occlusion. As a comparison, we propose another computational model that utilizes object parts/subparts in a compositional manner to build robustness to occlusion. This performs significantly better than CNN-based models on our task with error patterns similar to humans. These findings suggest that testing under extreme occlusion can better reveal the robustness of visual recognition, and that the principle of composition can encourage such robustness.

cs.CV↗