Yang Feng
Yang Feng
Scroll
Professor of Biostatistics
-
Professional overview
-
Yang Feng is a Professor and Ph.D. Program Director of Biostatistics in the School of Global Public Health and an affiliate faculty in the Center for Data Science at New York University. He obtained his Ph.D. in Operations Research at Princeton University in 2010.
Feng's research interests encompass the theoretical and methodological aspects of machine learning, high-dimensional statistics, social network models, and nonparametric statistics, leading to a wealth of practical applications, including Alzheimer's disease, cancer classification, and electronic health records. His research has been funded by multiple grants from the National Institutes of Health (NIH) and the National Science Foundation (NSF), notably the NSF CAREER Award.
He is currently an Associate Editor for the Journal of the American Statistical Association (JASA), the Journal of Business & Economic Statistics (JBES), Journal of Computational & Graphical Statistics (JCGS), and the Annals of Applied Statistics (AoAS). His professional recognitions include being named a fellow of the American Statistical Association (ASA) and the Institute of Mathematical Statistics (IMS), as well as an elected member of the International Statistical Institute (ISI).
Please visit Dr. Yang Feng's website and Google Scholar page from more information.
-
Education
-
B.S. in Mathematics, University of Science and Technology of China, Hefei, ChinaPh.D. in Operations Research, Princeton University, Princeton, NJ
-
Areas of research and study
-
BioinformaticsBiostatisticsHigh-dimensional data analysis/integrationMachine learningModeling Social and Behavioral DynamicsNonparametric statistics
-
Publications
Publications
Consistent Estimation of the Number of Communities in Non-uniform Hypergraph Model
AbstractShang, Z., Zhang, Z., & Feng, Y. (n.d.).Publication year
2025Journal title
StatVolume
14Issue
2AbstractWe propose an algorithm based on cross-validation to estimate the number of communities in a general non-uniform hypergraph model. The algorithm involves a three-step process. Initially, it randomly divides the set of hyperedges into a training set and a testing set. Subsequently, for each candidate number of communities, we construct a spectral estimation of community labels and least square estimation of the hyperedge probabilities based on the training set. The final step involves the computation of cross-validation scores using the testing set. The proposed algorithm is shown to be consistent when the number of vertices tends to infinity.Machine collaboration
AbstractLiu, Q., & Feng, Y. (n.d.).Publication year
2024Journal title
StatVolume
13Issue
1AbstractWe propose a new ensemble framework for supervised learning, called machine collaboration (MaC), using a collection of possibly heterogeneous base learning methods (hereafter, base machines) for prediction tasks. Unlike bagging/stacking (a parallel and independent framework) and boosting (a sequential and top-down framework), MaC is a type of circular and recursive learning framework. The circular and recursive nature helps the base machines to transfer information circularly and update their structures and parameters accordingly. The theoretical result on the risk bound of the estimator from MaC reveals that the circular and recursive feature can help MaC reduce risk via a parsimonious ensemble. We conduct extensive experiments on MaC using both simulated data and 119 benchmark real datasets. The results demonstrate that in most cases, MaC performs significantly better than several other state-of-the-art methods, including classification and regression trees, neural networks, stacking, and boosting.Machine collaboration
AbstractFeng, Y., Liu, Q., & Feng, Y. (n.d.).Publication year
2024Journal title
StatVolume
13Issue
1Page(s)
e661Abstract~Multi-label Random Subspace Ensemble Classification
AbstractBi, F., Zhu, J., & Feng, Y. (n.d.).Publication year
2024Journal title
Journal of Computational and Graphical StatisticsAbstractIn this work, we develop a new ensemble learning framework, multi-label Random Subspace Ensemble (mRaSE), for multi-label classification. Given a base classifier (e.g., multinomial logistic regression, classification tree, K-nearest neighbors), mRaSE works by first randomly sampling a collection of subspaces, then choosing the best ones that achieve the minimum cross-validation errors and, finally, aggregating the chosen weak learners. In addition to its superior prediction performance, mRaSE also provides a model-free feature ranking depending on the given base classifier. An iterative version of mRaSE is also developed to further improve the performance. A model-free extension is pursued on the iterative version, leading to the so-called Super mRaSE, which accepts a collection of base classifiers as input to the algorithm. We show the proposed algorithms compared favorably with the state-of-the-art classification algorithm including random forest and deep neural network, via extensive simulation studies and two real data applications. The new algorithms are implemented in an updated version of the R package RaSEn.Neyman-Pearson Multi-Class Classification via Cost-Sensitive Learning
AbstractTian, Y., & Feng, Y. (n.d.).Publication year
2024Journal title
Journal of the American Statistical AssociationAbstractMost existing classification methods aim to minimize the overall misclassification error rate. However, in applications such as loan default prediction, different types of errors can have varying consequences. To address this asymmetry issue, two popular paradigms have been developed: the Neyman-Pearson (NP) paradigm and the cost-sensitive (CS) paradigm. Previous studies on the NP paradigm have primarily focused on the binary case, while the multi-class NP problem poses a greater challenge due to its unknown feasibility. In this work, we tackle the multi-class NP problem by establishing a connection with the CS problem via strong duality and propose two algorithms. We extend the concept of NP oracle inequalities, crucial in binary classifications, to NP oracle properties in the multi-class context. Our algorithms satisfy these NP oracle properties under certain conditions. Furthermore, we develop practical algorithms to assess the feasibility and strong duality in multi-class NP problems, which can offer practitioners the landscape of a multi-class NP problem with various target error levels. Simulations and real data studies validate the effectiveness of our algorithms. To our knowledge, this is the first study to address the multi-class NP problem with theoretical guarantees. The proposed algorithms have been implemented in the R package npcs, which is available on CRAN. Supplementary materials for this article are available online, including a standardized description of the materials available for reproducing the work.Omics feature selection with the extended SIS R package : identification of a body mass index epigenetic multimarker in the Strong Heart Study
AbstractDomingo-Relloso, A., Feng, Y., Rodriguez-Hernandez, Z., Haack, K., Cole, S. A., Navas-Acien, A., Tellez-Plaza, M., & Bermudez, J. D. (n.d.).Publication year
2024Journal title
American Journal of EpidemiologyVolume
193Issue
7Page(s)
1010-1018AbstractThe statistical analysis of omics data poses a great computational challenge given their ultra–high-dimensional nature and frequent between-features correlation. In this work, we extended the iterative sure independence screening (ISIS) algorithm by pairing ISIS with elastic-net (Enet) and 2 versions of adaptive elastic-net (adaptive elastic-net (AEnet) and multistep adaptive elastic-net (MSAEnet)) to efficiently improve feature selection and effect estimation in omics research. We subsequently used genome-wide human blood DNA methylation data from American Indian participants in the Strong Heart Study (n = 2235 participants; measured in 1989-1991) to compare the performance (predictive accuracy, coefficient estimation, and computational efficiency) of ISIS-paired regularization methods with that of a bayesian shrinkage and traditional linear regression to identify an epigenomic multimarker of body mass index (BMI). ISIS-AEnet outperformed the other methods in prediction. In biological pathway enrichment analysis of genes annotated to BMI-related differentially methylated positions, ISIS-AEnet captured most of the enriched pathways in common for at least 2 of all the evaluated methods. ISIS-AEnet can favor biological discovery because it identifies the most robust biological pathways while achieving an optimal balance between bias and efficient feature selection. In the extended SIS R package, we also implemented ISIS paired with Cox and logistic regression for time-to-event and binary endpoints, respectively, and a bootstrap approach for the estimation of regression coefficients.Omics feature selection with the extended SIS R package: identification of a body mass index epigenetic multi-marker in the Strong Heart Study
AbstractFeng, Y., Domingo-Relloso, A., Feng, Y., Rodriguez-Hernandez, Z., Haack, K., Cole, S. A., Navas-Acien, A., Tellez-Plaza, M., & Bermudez, J. D. (n.d.).Publication year
2024Journal title
American Journal of EpidemiologyPage(s)
kwae006Abstract~Prognostic value of DNA methylation subclassification, aneuploidy, and CDKN2A/B homozygous deletion in predicting clinical outcome of IDH mutant astrocytomas
AbstractGalbraith, K., Garcia, M., Wei, S., Chen, A., Schroff, C., Serrano, J., Pacione, D., Placantonakis, D. G., William, C. M., Faustin, A., Zagzag, D., Barbaro, M., Del Pilar Guillermo Prieto Eibl, M., Shirahata, M., Reuss, D., Tran, Q. T., Alom, Z., von Deimling, A., Orr, B. A., … Snuderl, M. (n.d.).Publication year
2024Journal title
Neuro-OncologyVolume
26Issue
6Page(s)
1042-1051AbstractBackground. Isocitrate dehydrogenase (IDH) mutant astrocytoma grading, until recently, has been entirely based on morphology. The 5th edition of the Central Nervous System World Health Organization (WHO) introduces CDKN2A/B homozygous deletion as a biomarker of grade 4. We sought to investigate the prognostic impact of DNA methylation-derived molecular biomarkers for IDH mutant astrocytoma. Methods. We analyzed 98 IDH mutant astrocytomas diagnosed at NYU Langone Health between 2014 and 2022. We reviewed DNA methylation subclass, CDKN2A/B homozygous deletion, and ploidy and correlated molecular biomarkers with histological grade, progression free (PFS), and overall (OS) survival. Findings were confirmed using 2 independent validation cohorts. Results. There was no significant difference in OS or PFS when stratified by histologic WHO grade alone, copy number complexity, or extent of resection. OS was significantly different when patients were stratified either by CDKN2A/B homozygous deletion or by DNA methylation subclass (P value = .0286 and .0016, respectively). None of the molecular biomarkers were associated with significantly better PFS, although DNA methylation classification showed a trend (P value = .0534). Conclusions. The current WHO recognized grading criteria for IDH mutant astrocytomas show limited prognostic value. Stratification based on DNA methylation shows superior prognostic value for OS.Prognostic value of DNA methylation subclassification, aneuploidy, and CDKN2A/B homozygous deletion in predicting clinical outcome of IDH mutant astrocytomas
AbstractFeng, Y., Galbraith, K., Garcia, M., Wei, S., Chen, A., Schroff, C., Serrano, J., Pacione, D., Placantonakis, D. G., William, C. M., Faustin, A., & others. (n.d.).Publication year
2024Journal title
Neuro-OncologyPage(s)
noae009Abstract~Racial distribution of molecularly classified brain tumors
AbstractFang, C. S., Wang, W., Schroff, C., Movahed-Ezazi, M., Vasudevaraja, V., Serrano, J., Sulman, E. P., Golfinos, J. G., Orringer, D., Galbraith, K., Feng, Y., & Snuderl, M. (n.d.).Publication year
2024Journal title
Neuro-Oncology AdvancesVolume
6Issue
1AbstractBackground. In many cancers, specific subtypes are more prevalent in specific racial backgrounds. However, little is known about the racial distribution of specific molecular types of brain tumors. Public data repositories lack data on many brain tumor subtypes as well as diagnostic annotation using the current World Health Organization classification. A better understanding of the prevalence of brain tumors in different racial backgrounds may provide insight into tumor predisposition and development, and improve prevention. Methods. We retrospectively analyzed the racial distribution of 1709 primary brain tumors classified by their methylation profiles using clinically validated whole genome DNA methylation. Self-reported race was obtained from medical records. Our cohort included 82% White, 10% Black, and 8% Asian patients with 74% of patients reporting their race. Results. There was a significant difference in the racial distribution of specific types of brain tumors. Blacks were overrepresented in pituitary adenomas (35%, P < .001), with the largest proportion of FSH/LH subtype. Whites were underrepresented at 47% of all pituitary adenoma patients (P < .001). Glioblastoma (GBM) IDH wild-type showed an enrichment of Whites, at 90% (P < .001), and a significantly smaller percentage of Blacks, at 3% (P < .001). Conclusions. Molecularly classified brain tumor groups and subgroups show different distributions among the three main racial backgrounds suggesting the contribution of race to brain tumor development.Towards the Theory of Unsupervised Federated Learning : Non-asymptotic Analysis of Federated EM Algorithms
AbstractTian, Y., Weng, H., & Feng, Y. (n.d.).Publication year
2024Journal title
Proceedings of Machine Learning ResearchVolume
235Page(s)
48226-48279AbstractWhile supervised federated learning approaches have enjoyed significant success, the domain of unsupervised federated learning remains relatively underexplored. Several federated EM algorithms have gained popularity in practice, however, their theoretical foundations are often lacking. In this paper, we first introduce a federated gradient EM algorithm (FedGrEM) designed for the unsupervised learning of mixture models, which supplements the existing federated EM algorithms by considering task heterogeneity and potential adversarial attacks. We present a comprehensive finite-sample theory that holds for general mixture models, then apply this general theory on specific statistical models to characterize the explicit estimation error of model parameters and mixture proportions. Our theory elucidates when and how FedGrEM outperforms local single-task learning with insights extending to existing federated EM algorithms. This bridges the gap between their practical success and theoretical understanding. Our numerical results validate our theory, and demonstrate FedGrEM’s superiority over existing unsupervised federated learning benchmarks.ℓ1-Penalized Multinomial Regression : Estimation, Inference, and Prediction, With an Application to Risk Factor Identification for Different Dementia Subtypes
AbstractTian, Y., Rusinek, H., Masurkar, A. V., & Feng, Y. (n.d.).Publication year
2024Journal title
Statistics in MedicineVolume
43Issue
30Page(s)
5711-5747AbstractHigh-dimensional multinomial regression models are very useful in practice but have received less research attention than logistic regression models, especially from the perspective of statistical inference. In this work, we analyze the estimation and prediction error of the contrast-based (Formula presented.) -penalized multinomial regression model and extend the debiasing method to the multinomial case, providing a valid confidence interval for each coefficient and (Formula presented.) value of the individual hypothesis test. We also examine cases of model misspecification and non-identically distributed data to demonstrate the robustness of our method when some assumptions are violated. We apply the debiasing method to identify important predictors in the progression into dementia of different subtypes. Results from extensive simulations show the superiority of the debiasing method compared to other inference methods.A flexible quasi-likelihood model for microbiome abundance count data
AbstractFeng, Y., Shi, Y., Li, H., Wang, C., Chen, J., Jiang, H., Shih, Y.-C. T., Zhang, H., Song, Y., Feng, Y., & Liu, L. (n.d.).Publication year
2023Journal title
Statistics in MedicineVolume
42Issue
25Page(s)
4632--4643Abstract~Comments on: Statistical inference and large-scale multiple testing for high-dimensional regression models
AbstractFeng, Y., Tian, Y. e., & Feng, Y. (n.d.).Publication year
2023Journal title
TestVolume
32Issue
4Page(s)
1172--1176Abstract~DDAC-SpAM: A Distributed Algorithm for Fitting High-dimensional Sparse Additive Models with Feature Division and Decorrelation
AbstractFeng, Y., He, Y., Wu, R., Zhou, Y., & Feng, Y. (n.d.).Publication year
2023Journal title
Journal of the American Statistical AssociationPage(s)
1--12Abstract~Design-Based Causal Inference with Missing Outcomes: Missingness Mechanisms, Imputation-Assisted Randomization Tests, and Covariate Adjustment
AbstractFeng, Y., Heng, S., Zhang, J., & Feng, Y. (n.d.).Publication year
2023Journal title
arXiv preprint arXiv:2310.18556Abstract~Learning from Similar Linear Representations: Adaptivity, Minimaxity, and Robustness
AbstractFeng, Y., Tian, Y. e., Gu, Y., & Feng, Y. (n.d.).Publication year
2023Journal title
arXiv preprint arXiv:2303.17765Abstract~PCABM: Pairwise Covariates-Adjusted Block Model for Community Detection
AbstractFeng, Y., Huang, S., Sun, J., & Feng, Y. (n.d.).Publication year
2023Journal title
Journal of the American Statistical AssociationPage(s)
1--13Abstract~Semiparametric Modeling and Analysis for Longitudinal Network Data
AbstractFeng, Y., He, Y., Sun, J., Tian, Y., Ying, Z., & Feng, Y. (n.d.).Publication year
2023Journal title
arXiv preprint arXiv:2308.12227Abstract~Simulation of New York City’s Ventilator Allocation Guideline During the Spring 2020 COVID-19 Surge
AbstractFeng, Y., Walsh, B. C., Zhu, J., Feng, Y., Berkowitz, K. A., Betensky, R. A., Nunnally, M. E., & Pradhan, D. R. (n.d.).Publication year
2023Journal title
JAMA network openVolume
6Issue
10Page(s)
e2336736--e2336736Abstract~Transfer learning under high-dimensional generalized linear models
AbstractFeng, Y., Tian, Y. e., & Feng, Y. (n.d.).Publication year
2023Journal title
Journal of the American Statistical AssociationVolume
118Issue
544Page(s)
2684--2697Abstract~Unsupervised Federated Learning: A Federated Gradient EM Algorithm for Heterogeneous Mixture Models with Robustness against Adversarial Attacks
AbstractFeng, Y., Tian, Y. e., Weng, H., & Feng, Y. (n.d.).Publication year
2023Journal title
arXiv preprint arXiv:2310.15330Abstract~Variable selection for high-dimensional generalized linear model with block-missing data
AbstractFeng, Y., He, Y., Feng, Y., & Song, X. (n.d.).Publication year
2023Journal title
Scandinavian Journal of StatisticsVolume
50Issue
3Page(s)
1279--1297Abstract~A likelihood-ratio type test for stochastic block models with bounded degrees
AbstractFeng, Y., Yuan, M., Feng, Y., & Shang, Z. (n.d.).Publication year
2022Journal title
Journal of Statistical Planning and InferenceVolume
219Page(s)
98--119Abstract~Association of hyperglycemia and molecular subclass on survival in IDH-wildtype glioblastoma
AbstractFeng, Y., Liu, E. K., Vasudevaraja, V., Sviderskiy, V. O., Feng, Y., Tran, I., Serrano, J., Cordova, C., Kurz, S. C., Golfinos, J. G., Sulman, E. P., & others. (n.d.).Publication year
2022Journal title
Neuro-Oncology AdvancesVolume
4Issue
1Page(s)
vdac163Abstract~