arXiv subjects

Tiffany M. Tang

Publications and source records attributed to Tiffany M. Tang.

12 recordsLinked to original sources

arXiv2026

Interpretable Network-assisted Random Forest+

Machine learning algorithms often assume that training samples are independent. When data points are connected by a network, the induced dependency between samples is both a challenge, reducing effective sample size, and an opportunity to improve prediction by leveraging information from network neighbors. Multiple methods taking advantage of this opportunity are available, but many, including graph neural networks, are not easily interpretable, limiting their usefulness for understanding how models make predictions. Others, such as network-assisted linear regression, are interpretable but often yield worse prediction performance. We bridge this gap by proposing a family of flexible network-assisted models built upon a generalization of random forests (RF+), which achieves highly-competitive prediction accuracy and can be understood through intrinsic interpretability measures, derived directly from the model parameters and structure. In particular, we develop a suite of interpretation tools that enable researchers to both identify important features that drive model predictions and quantify the importance of the network contribution to prediction. Importantly, we provide global and local feature importances as well as sample influence measures to assess the impact of individual observations. This suite of tools broadens the scope and applicability of network-assisted machine learning for high-impact problems where interpretability and transparency are essential.

stat.ML↗

arXiv2026

Distilling heterogeneous treatment effects: Stable subgroup estimation in causal inference

Recent methodological developments have introduced new black-box approaches to better estimate heterogeneous treatment effects; however, these methods fall short of providing interpretable characterizations of the underlying individuals who may be most at risk or benefit most from receiving the treatment, thereby limiting their practical utility. In this work, we introduce causal distillation trees (CDT) to estimate interpretable subgroups. CDT allows researchers to fit any machine learning model to estimate heterogeneous treatment effects, and then leverages a simple, second-stage tree-based model to "distill" the estimated treatment effects into meaningful subgroups. As a result, CDT inherits the improvements in predictive performance from black-box machine learning models while preserving the interpretability of a simple decision tree. We derive theoretical guarantees for the consistency of the estimated subgroups using CDT, and introduce stability-driven diagnostics for researchers to evaluate the quality of the estimated subgroups. We illustrate our proposed method on a randomized controlled trial of antiretroviral treatment for HIV from the AIDS Clinical Trials Group Study 175 and show that CDT out-performs state-of-the-art approaches in constructing stable, clinically relevant subgroups.

stat.ME↗

arXiv2026

Cluster Analysis with Resampling for Validation and Exploration (CARVE)

Clustering is widely used across the sciences as the foundation for downstream data-driven scientific discoveries. However, clustering results are highly sensitive to the choice of algorithm, preprocessing, and the number of clusters $k$, producing scientific claims that are often not reproducible. The current state of the art for validating clustering solutions consists of clustering validation indices (CVIs) such as Silhouette, Davies-Bouldin, and Calinski-Harabasz, which rely on geometric assumptions that break down on the heavy-tailed, high-dimensional, and nonlinearly structured data encountered in biomedical research. Resampling-based alternatives - grounded in the ideas of clustering stability and generalizability - have been proposed but remain scattered across specialized tools with no unified, accessible software. We fill this gap with CARVE (Cluster Analysis with Resampling for Validation and Exploration), an open-source Python and R package that jointly evaluates multiple clustering algorithms and hyperparameters, returning stability and generalizability diagnostics at the global, cluster, and sample level together with principled selection rules and consensus-based cluster labels. Across six synthetic benchmarks CARVE consistently recovers near-optimal clusterings where classical indices degrade substantially. On experimental genomics and proteomics data sets, CARVE recovers finer biological structure when classical CVIs collapse entirely. CARVE is available with a scikit-learn-compatible Python API and an analogous R interface compatible with Seurat workflows.

stat.ME↗

arXiv2026

Local MDI+: Local Feature Importances for Tree-Based Models

Tree-based ensembles such as random forests remain the go-to for tabular data over deep learning models due to their prediction performance and computational efficiency. These advantages have led to their widespread deployment in high-stakes domains, where interpretability is essential for ensuring trustworthy predictions. This has motivated the development of popular local feature importance methods such as LIME and TreeSHAP. However, these approaches rely on approximations that ignore the model's internal structure and instead depend on potentially unstable perturbations. These issues are addressed in the global setting by MDI+, a global feature importance method which combines tree-based and linear feature importances by exploiting an equivalence between decision trees and least squares on a transformed node basis. However, the global MDI+ scores are not able to explain predictions when faced with heterogeneous individual characteristics. To address this gap, we propose Local MDI+ (LMDI+), a novel extension of the MDI+ framework that quantifies feature importances for each particular sample. Across twelve real-world benchmark datasets, LMDI+ outperforms existing baselines at identifying instance-specific predictive features, yielding an average 10% improvement in predictive performance when using only the selected features. It further demonstrates greater stability by consistently producing similar instance-level feature importance rankings across repeated model fits with different random seeds. Ablation experiments show that each component of LMDI+ contributes to these gains, and that the improvements extend beyond random forests to gradient boosting models. Finally, we show that LMDI+ enables local interpretability use cases by identifying closely matched counterfactuals for each classification benchmark and discovering homogeneous subgroups in a housing dataset case study.

cs.LG↗

arXiv2026

Estimating Consensus Ideal Points Using Multi-Source Data

In the advent of big data and machine learning, researchers now have a wealth of congressional candidate ideal point estimates at their disposal for theory testing. Weak relationships raise questions about the extent to which they capture a shared quantity -- rather than idiosyncratic, domain-specific factors -- yet different measures are used interchangeably in most substantive analyses. Moreover, questions central to the study of American politics implicate relationships between candidate ideal points and other variables derived from the same data sources, introducing endogeneity. We propose a method, consensus multidimensional scaling (CoMDS), which better aligns with how applied scholars use ideal points in practice. CoMDS captures the shared, stable associations of a set of underlying ideal point estimates and can be interpreted as their common spatial representation. We illustrate the utility of our approach for assessing relationships within domains of existing measures and provide a suite of diagnostic tools to aid in practical usage.

stat.AP↗

arXiv2025

Consensus dimension reduction via multi-view learning

A plethora of dimension reduction methods have been developed to visualize high-dimensional data in low dimensions. However, different dimension reduction methods often output different and possibly conflicting visualizations of the same data. This problem is further exacerbated by the choice of hyperparameters, which may substantially impact the resulting visualization. To obtain a more robust and trustworthy dimension reduction output, we advocate for a consensus approach, which summarizes multiple visualizations into a single consensus dimension reduction visualization. Here, we leverage ideas from multi-view learning in order to identify the patterns that are most stable or shared across the many different dimension reduction visualizations, or views, and subsequently visualize this shared structure in a single low-dimensional plot. We demonstrate that this consensus visualization effectively identifies and preserves the shared low-dimensional data structure through both simulated and real-world case studies. We further highlight our method's robustness to the choice of dimension reduction method and hyperparameters -- a highly-desirable property when working towards trustworthy and reproducible data science.

stat.ME↗

arXiv2025

Estimating Fe and Mg Abundances in the Milky Way Dwarf Galaxies Using Subaru/HSC and DEIMOS

We investigate the chemical abundance distributions of the Fornax, Sculptor, Ursa Minor, and Draco dwarf galaxies using Subaru/HSC photometric data. The HSC dataset, which includes broadband g and i filters and the narrowband NB515 filter, offers sensitivity to iron and magnesium abundances as well as surface gravity, enabling the identification of giant stars and foreground dwarfs. For analysis, we selected a total of 6713 giant candidates using a Random Forest regressor trained on medium-resolution (R ~ 6000) Keck/DEIMOS spectroscopic data. Our analysis reveals the extent of radial metallicity gradients in the galaxies. Such trends, not detectable in earlier studies, are now captured owing to the substantially enlarged sample size and areal coverage provided by the HSC data. These results are also consistent with chemical abundance patterns previously observed in the central regions through spectroscopic studies. Furthermore, we infer that Fornax underwent extended star formation, whereas Sculptor formed both metal-poor and metal-rich stars over a shorter time. Ursa Minor and Draco appear to have experienced brief, intense star formation episodes leading to nearly extinguished star formation. This study underscores the critical role of the expanded HSC dataset in revealing chemical gradients that were previously inaccessible. Future work incorporating additional spectra of metal-poor stars and age-sensitive isochrone modeling will enable more accurate maps of chemical abundance distributions.

astro-ph.GA↗

arXiv2025

Unsupervised Machine Learning for Scientific Discovery: Workflow and Best Practices

Unsupervised machine learning is widely used to mine large, unlabeled datasets to make data-driven discoveries in critical domains such as climate science, biomedicine, astronomy, chemistry, and more. However, despite its widespread utilization, there is a lack of standardization in unsupervised learning workflows for making reliable and reproducible scientific discoveries. In this paper, we present a structured workflow for using unsupervised learning techniques in science. We highlight and discuss best practices starting with formulating validatable scientific questions, conducting robust data preparation and exploration, using a range of modeling techniques, performing rigorous validation by evaluating the stability and generalizability of unsupervised learning conclusions, and promoting effective communication and documentation of results to ensure reproducible scientific discoveries. To illustrate our proposed workflow, we present a case study from astronomy, seeking to refine globular clusters of Milky Way stars based upon their chemical composition. Our case study highlights the importance of validation and illustrates how the benefits of a carefully-designed workflow for unsupervised learning can advance scientific discovery.

cs.LG↗

arXiv2025

Integrating Random Forests and Generalized Linear Models for Improved Accuracy and Interpretability

Random forests (RFs) are among the most popular supervised learning algorithms due to their nonlinear flexibility and ease-of-use. However, as black box models, they can only be interpreted via algorithmically-defined feature importance methods, such as Mean Decrease in Impurity (MDI), which have been observed to be highly unstable and have ambiguous scientific meaning. Furthermore, they can perform poorly in the presence of smooth or additive structure. To address this, we reinterpret decision trees and MDI as linear regression and $R^2$ values, respectively, with respect to engineered features associated with the tree's decision splits. This allows us to combine the respective strengths of RFs and generalized linear models in a framework called RF+, which also yields an improved feature importance method we call MDI+. Through extensive data-inspired simulations and real-world datasets, we show that RF+ improves prediction accuracy over RFs and that MDI+ outperforms popular feature importance measures in identifying signal features, often yielding more than a 10% improvement over its closest competitor. In case studies on drug response prediction and breast cancer subtyping, we further show that MDI+ extracts well-established genes with significantly greater stability compared to existing feature importance measures.

stat.ME↗

arXiv2021

Integrated Principal Components Analysis

Data integration, or the strategic analysis of multiple sources of data simultaneously, can often lead to discoveries that may be hidden in individualistic analyses of a single data source. We develop a new unsupervised data integration method named Integrated Principal Components Analysis (iPCA), which is a model-based generalization of PCA and serves as a practical tool to find and visualize common patterns that occur in multiple data sets. The key idea driving iPCA is the matrix-variate normal model, whose Kronecker product covariance structure captures both individual patterns within each data set and joint patterns shared by multiple data sets. Building upon this model, we develop several penalized (sparse and non-sparse) covariance estimators for iPCA, and using geodesic convexity, we prove that our non-sparse iPCA estimator converges to the global solution of a non-convex problem. We also demonstrate the practical advantages of iPCA through extensive simulations and a case study application to integrative genomics for Alzheimer's disease. In particular, we show that the joint patterns extracted via iPCA are highly predictive of a patient's cognition and Alzheimer's diagnosis.

stat.ME↗

arXiv2020

A stability-driven protocol for drug response interpretable prediction (staDRIP)

Modern cancer -omics and pharmacological data hold great promise in precision cancer medicine for developing individualized patient treatments. However, high heterogeneity and noise in such data pose challenges for predicting the response of cancer cell lines to therapeutic drugs accurately. As a result, arbitrary human judgment calls are rampant throughout the predictive modeling pipeline. In this work, we develop a transparent stability-driven pipeline for drug response interpretable predictions, or staDRIP, which builds upon the PCS framework for veridical data science (Yu and Kumbier, 2020) and mitigates the impact of human judgment calls. Here we use the PCS framework for the first time in cancer research to extract proteins and genes that are important in predicting the drug responses and stable across appropriate data and model perturbations. Out of the 24 most stable proteins we identified using data from the Cancer Cell Line Encyclopedia (CCLE), 18 have been associated with the drug response or identified as a known or possible drug target in previous literature, demonstrating the utility of our stability-driven pipeline for knowledge discovery in cancer drug response prediction modeling.

q-bio.QM↗

arXiv2020

Feature Selection for Data Integration with Mixed Multi-view Data

Data integration methods that analyze multiple sources of data simultaneously can often provide more holistic insights than can separate inquiries of each data source. Motivated by the advantages of data integration in the era of "big data", we investigate feature selection for high-dimensional multi-view data with mixed data types (e.g. continuous, binary, count-valued). This heterogeneity of multi-view data poses numerous challenges for existing feature selection methods. However, after critically examining these issues through empirical and theoretically-guided lenses, we develop a practical solution, the Block Randomized Adaptive Iterative Lasso (B-RAIL), which combines the strengths of the randomized Lasso, adaptive weighting schemes, and stability selection. B-RAIL serves as a versatile data integration method for sparse regression and graph selection, and we demonstrate the effectiveness of B-RAIL through extensive simulations and a case study to infer the ovarian cancer gene regulatory network. In this case study, B-RAIL successfully identifies well-known biomarkers associated with ovarian cancer and hints at novel candidates for future ovarian cancer research.

stat.ME↗