Journal Articles
39. Dynamic treatment effects: high-dimensional inference under model misspecification | 2021+
New Loss functions for efficient inference over time
(with Yuqian Zhang and Weijie Ji)
38. High dimensional inference for dynamic treatment effects | 2021+
Dynamic Causal inference with Lasso
(with Weijie Ji and Yuqian Zhang)
37. Doubly robust semi-supervised inference for the mean:
selection bias under MAR labeling with decaying overlap | 2021+
Causal inference with uniformly decaying propensity
(with Yuqian Zhang and Abhishek Chakrabortty)
36. Comments on Leo Breiman's paper
'Statistical Modeling: The Two Cultures' (Statistical Science, 2001, 16(3), 199-231) | Observational Studies, to appear
Double robustness and machine learning: friends or foe.
(with Yinchu Zhu)
35. Dynamic covariate balancing: estimating treatment effects over time | 2021+
Balancing covariates when observations are over time and high dimensional
(with Davide Viviano )
34. Learning to combat the noisy labels via classification margins | 2021+
New Margins via early learning (MARVEL) classifier as a robust neural network
(with Jason Z. Lin )
33. Deep Hazard: neural network for time-varying risks | 2020+
Architecture for learning survival outcomes with observations
that are censored and vary with time
(with Denise Rava)
32. Fair policy targeting | 2020+
Counterfactual Envy-Free fairness
via Pareto optimality for fairness
(with Davide Viviano)
31. Rejoinder of "A tuning free robust and efficient approach to high-dimensional regression" | JASA:T&M, 115 (532), 1726-1729, (2020)
Variable selection without tuning that is robust as well as efficient
(with Lan Wang, Runze Li, Bo Peng and Yunan Wu )
30. A tuning free robust and efficient approach to high-dimensional regression | JASA:T&M, with discussion, 115 (532), 1700-1714, (2020)
Variable selection without tuning that is robust as well as efficient
(with Lan Wang, Runze Li, Bo Peng and Yunan Wu )
29. Causal quantile learner: robust structural equations | 2020+
Proposed quantile invariance for learning cause-effect
relationship between many variables with
interventional/observational data
(with Denise Rava)
28. Minimax semiparametric learning with approximate sparsity | 2019+
Root-n optimality of double robustness under approximate sparsity
that illustrates differences from classical semiparametrics
(with Victor Chernozhukov, Whitney K. Newey and Yinchu Zhu)
27. Sparsity double robust inference of average treatment effects | 2019+
Proposed new balancing via moment-targeting for average
treatment effect estimation and inference
(with Stefan Wager and Yinchu Zhu)
26. Synthetic learner: model-free inference on treatments over time | JOE, revision
Inference for counterfactuals using machine learning
where the true model may not be captured in the dictionary
(with Davide Viviano)
25. High-dimensional semi-supervised learning: in search of optimal inference of the mean | Biometrika, minor revision
Semi-supervised inference of the responses mean when covariates
are high-dimensional and model is not necessarily correctly specified
(with Yuqian Zhang)
24. Estimating treatment effects under additive hazards models with high-dimensional confounding | JASA:T&M, forthcoming
Double robust treatment effects in additive hazards in the presence of right censoring
(with Jue Hou and Ronghui Xu)
23. Minimax rates and adaptivity of tests in high-dimensional linear models with non-sparse structures | AOS, to appear
New minimax optimality results regarding confidence intervals -- no sparsity restrictions needed
(with Jianqing Fan and Yinchu Zhu )
22. Confidence intervals for high-dimensional Cox models | Statistica Sinica, 31(1), 243-267, (2021)
Assumption-lean asymptotic theory for the inference in the Cox model
(with Yi Yu and Richard J. Samworth )
21. Detangling robustness in high-dimensions: composite vs model-averaged estimation | EJS, 14(2), 2551-2599, (2020)
Illustrated that composite estimation has benefits over model-averaged
estimation in high-dimensions with sparsity
(with Jing Zhou and Gerda Claeskens)
20. Censored quantile regression forest | AIStats, 108, 2109-2119, (2020)
Censored quantile random forest estimates
quantiles of censored responses in a regression setting non-parametrically.
(with Alexander Hanbo Li )
19. Asymptotic theory of rank estimation for high-dimensional accelerated failure time models | AOS, revision
Finite-sample estimation properties of new regularizer for AFT models
(with Lan Wang )
18. High-dimensional classification with errors in variables using high-confidence sets | 2018+
Classifiers for noisy data with possibly non-sparse classification boundaries
(with Emre Barut, Jianqing Fan and Jiancheng Jiang )
17. Testing fixed effects in high-dimensional misspecified linear mixed models | JASA:T&M, 115(532), 1835-1850, (2020)
Estimators and tests adaptive to misspecification of random effects
(with Gerda Claeskens and Thomas Gueuning )
16. Inference under Fine-Gray competing risks model with high-dimensional covariates | EJS, 13(2), 4449-4507, (2019)
Regularization and testing for Fine-Gray model with many more parameters than samples
(with Jue Hou and Ronghui Xu )
15. Breaking the curse of dimensionality in high-dimensions | JMLR, revision
Tests of multivariate parameters in high-dimensional setting
that does not rely on the sparsity of the underlying linear model
(with Yinchu Zhu)
14. A projection pursuit framework for testing general high-dimensional hypothesis | 2017+
New projection estimator and test statistic based on the contrast of two competing estimators
(with Yinchu Zhu)
13. Two-sample testing in high-dimensional and dense models | 2016+
New algorithm TIERS for distinguishing between coefficients of two
high-dimensional regressions
(with Yinchu Zhu)
12. Uniform inference for high-dimensional quantile process: linear testing and regression rank scores | AOS, revision
Uniform bahadur representation, dual problems and uniform multivariate tests
(with Mladen Kolar)
11. Generalized M-estimators for high-dimensional tobit I models | EJS, 13(1), 582-645, (2019)
Mallow's, Hill-Ryan's and Sweepe's one-step estimators for
left (fixed) censored data: Tobit I model and its variants
(with Jiaqi Guo)
10. Linear hypothesis testing in dense high-dimensional linear models | JASA:T&M, 113(524), 1583-1600, (2018)
New restructured regression method for testing hypothesis with
dense parameters and dense loadings in high-dimensions
(with Yinchu Zhu)
9. Significance testing in non-sparse high-dimensional linear models | EJS, 12(2), 3312-3364 (2018)
New method CorrT proposed that preserves Type I error and Type II error even in
dense (non-sparse) and ultra high-dimensional models
(with Yinchu Zhu)
8. Boosting in the presence of outliers: adaptive classification in the presence of outliers | JASA:T&M , 113(512), 660-674, (2018)
New boosting method ArchBoost that is robust to data or label perturbations
-- adversarial or not
(with Alexander Hanbo Li)
7. Comment on "High dimensional simultaneous inference via bootstrap"
by R. Dezeure, P. Buhlmann and C-H. Zhang | TEST, 26(4), 720-728 (2017)
Discussed residual bootstrap efficiency and proposed new residual bootstrap
for mixture of sparse and dense models
(with Yinchu Zhu)
6. Robustness in sparse high-dimensional models:
relative efficiency based on approximate message passing | EJS, 10(2), 3894-3944 (2016)
New AMP algorithm is proposed, RAMP, that is shown to be
efficient regardless of the error distribution
5. Randomized maximum contrast selection: subagging for large-scale regression | EJS, 10(1), 121-170, (2016)
Model selection for big data: naive selection of variables fails
whereas maximum contrast selection succeeds
4. Cultivating disaster donors using data analytics | Mgmt. Science, 62(3), 849-866, (2016)
Importance sampling and logistic regression in
divide and conquer setting
(with Ilya Ryzhov and Bin Han)
3. Structured estimation in non-parametric Cox model | EJS, 9(1), 492-534, (2015)
Estimation in misspecified high-dimensional Cox model
(with Rui Song)
2. Regularization for Cox's proportional hazard model with NP dimensionality | AOS, 39(6), 3092-3120, (2011)
Lasso and SCAD model selection properties for ultra high dimensional data
(with Jianqing Fan and Jiancheng Jiang)
1. Composite quasi-likelihood for high-dimensional variable selection | JRSSB, 73(3), 325-349, (2011)
Model selection robust and adaptive to the error distribution
(with Jianqing Fan and Weiwei Wang)
New Loss functions for efficient inference over time
(with Yuqian Zhang and Weijie Ji)
38. High dimensional inference for dynamic treatment effects | 2021+
Dynamic Causal inference with Lasso
(with Weijie Ji and Yuqian Zhang)
37. Doubly robust semi-supervised inference for the mean:
selection bias under MAR labeling with decaying overlap | 2021+
Causal inference with uniformly decaying propensity
(with Yuqian Zhang and Abhishek Chakrabortty)
36. Comments on Leo Breiman's paper
'Statistical Modeling: The Two Cultures' (Statistical Science, 2001, 16(3), 199-231) | Observational Studies, to appear
Double robustness and machine learning: friends or foe.
(with Yinchu Zhu)
35. Dynamic covariate balancing: estimating treatment effects over time | 2021+
Balancing covariates when observations are over time and high dimensional
(with Davide Viviano )
34. Learning to combat the noisy labels via classification margins | 2021+
New Margins via early learning (MARVEL) classifier as a robust neural network
(with Jason Z. Lin )
33. Deep Hazard: neural network for time-varying risks | 2020+
Architecture for learning survival outcomes with observations
that are censored and vary with time
(with Denise Rava)
32. Fair policy targeting | 2020+
Counterfactual Envy-Free fairness
via Pareto optimality for fairness
(with Davide Viviano)
31. Rejoinder of "A tuning free robust and efficient approach to high-dimensional regression" | JASA:T&M, 115 (532), 1726-1729, (2020)
Variable selection without tuning that is robust as well as efficient
(with Lan Wang, Runze Li, Bo Peng and Yunan Wu )
30. A tuning free robust and efficient approach to high-dimensional regression | JASA:T&M, with discussion, 115 (532), 1700-1714, (2020)
Variable selection without tuning that is robust as well as efficient
(with Lan Wang, Runze Li, Bo Peng and Yunan Wu )
29. Causal quantile learner: robust structural equations | 2020+
Proposed quantile invariance for learning cause-effect
relationship between many variables with
interventional/observational data
(with Denise Rava)
28. Minimax semiparametric learning with approximate sparsity | 2019+
Root-n optimality of double robustness under approximate sparsity
that illustrates differences from classical semiparametrics
(with Victor Chernozhukov, Whitney K. Newey and Yinchu Zhu)
27. Sparsity double robust inference of average treatment effects | 2019+
Proposed new balancing via moment-targeting for average
treatment effect estimation and inference
(with Stefan Wager and Yinchu Zhu)
26. Synthetic learner: model-free inference on treatments over time | JOE, revision
Inference for counterfactuals using machine learning
where the true model may not be captured in the dictionary
(with Davide Viviano)
25. High-dimensional semi-supervised learning: in search of optimal inference of the mean | Biometrika, minor revision
Semi-supervised inference of the responses mean when covariates
are high-dimensional and model is not necessarily correctly specified
(with Yuqian Zhang)
24. Estimating treatment effects under additive hazards models with high-dimensional confounding | JASA:T&M, forthcoming
Double robust treatment effects in additive hazards in the presence of right censoring
(with Jue Hou and Ronghui Xu)
23. Minimax rates and adaptivity of tests in high-dimensional linear models with non-sparse structures | AOS, to appear
New minimax optimality results regarding confidence intervals -- no sparsity restrictions needed
(with Jianqing Fan and Yinchu Zhu )
22. Confidence intervals for high-dimensional Cox models | Statistica Sinica, 31(1), 243-267, (2021)
Assumption-lean asymptotic theory for the inference in the Cox model
(with Yi Yu and Richard J. Samworth )
21. Detangling robustness in high-dimensions: composite vs model-averaged estimation | EJS, 14(2), 2551-2599, (2020)
Illustrated that composite estimation has benefits over model-averaged
estimation in high-dimensions with sparsity
(with Jing Zhou and Gerda Claeskens)
20. Censored quantile regression forest | AIStats, 108, 2109-2119, (2020)
Censored quantile random forest estimates
quantiles of censored responses in a regression setting non-parametrically.
(with Alexander Hanbo Li )
19. Asymptotic theory of rank estimation for high-dimensional accelerated failure time models | AOS, revision
Finite-sample estimation properties of new regularizer for AFT models
(with Lan Wang )
18. High-dimensional classification with errors in variables using high-confidence sets | 2018+
Classifiers for noisy data with possibly non-sparse classification boundaries
(with Emre Barut, Jianqing Fan and Jiancheng Jiang )
17. Testing fixed effects in high-dimensional misspecified linear mixed models | JASA:T&M, 115(532), 1835-1850, (2020)
Estimators and tests adaptive to misspecification of random effects
(with Gerda Claeskens and Thomas Gueuning )
16. Inference under Fine-Gray competing risks model with high-dimensional covariates | EJS, 13(2), 4449-4507, (2019)
Regularization and testing for Fine-Gray model with many more parameters than samples
(with Jue Hou and Ronghui Xu )
15. Breaking the curse of dimensionality in high-dimensions | JMLR, revision
Tests of multivariate parameters in high-dimensional setting
that does not rely on the sparsity of the underlying linear model
(with Yinchu Zhu)
14. A projection pursuit framework for testing general high-dimensional hypothesis | 2017+
New projection estimator and test statistic based on the contrast of two competing estimators
(with Yinchu Zhu)
13. Two-sample testing in high-dimensional and dense models | 2016+
New algorithm TIERS for distinguishing between coefficients of two
high-dimensional regressions
(with Yinchu Zhu)
12. Uniform inference for high-dimensional quantile process: linear testing and regression rank scores | AOS, revision
Uniform bahadur representation, dual problems and uniform multivariate tests
(with Mladen Kolar)
11. Generalized M-estimators for high-dimensional tobit I models | EJS, 13(1), 582-645, (2019)
Mallow's, Hill-Ryan's and Sweepe's one-step estimators for
left (fixed) censored data: Tobit I model and its variants
(with Jiaqi Guo)
10. Linear hypothesis testing in dense high-dimensional linear models | JASA:T&M, 113(524), 1583-1600, (2018)
New restructured regression method for testing hypothesis with
dense parameters and dense loadings in high-dimensions
(with Yinchu Zhu)
9. Significance testing in non-sparse high-dimensional linear models | EJS, 12(2), 3312-3364 (2018)
New method CorrT proposed that preserves Type I error and Type II error even in
dense (non-sparse) and ultra high-dimensional models
(with Yinchu Zhu)
8. Boosting in the presence of outliers: adaptive classification in the presence of outliers | JASA:T&M , 113(512), 660-674, (2018)
New boosting method ArchBoost that is robust to data or label perturbations
-- adversarial or not
(with Alexander Hanbo Li)
7. Comment on "High dimensional simultaneous inference via bootstrap"
by R. Dezeure, P. Buhlmann and C-H. Zhang | TEST, 26(4), 720-728 (2017)
Discussed residual bootstrap efficiency and proposed new residual bootstrap
for mixture of sparse and dense models
(with Yinchu Zhu)
6. Robustness in sparse high-dimensional models:
relative efficiency based on approximate message passing | EJS, 10(2), 3894-3944 (2016)
New AMP algorithm is proposed, RAMP, that is shown to be
efficient regardless of the error distribution
5. Randomized maximum contrast selection: subagging for large-scale regression | EJS, 10(1), 121-170, (2016)
Model selection for big data: naive selection of variables fails
whereas maximum contrast selection succeeds
4. Cultivating disaster donors using data analytics | Mgmt. Science, 62(3), 849-866, (2016)
Importance sampling and logistic regression in
divide and conquer setting
(with Ilya Ryzhov and Bin Han)
3. Structured estimation in non-parametric Cox model | EJS, 9(1), 492-534, (2015)
Estimation in misspecified high-dimensional Cox model
(with Rui Song)
2. Regularization for Cox's proportional hazard model with NP dimensionality | AOS, 39(6), 3092-3120, (2011)
Lasso and SCAD model selection properties for ultra high dimensional data
(with Jianqing Fan and Jiancheng Jiang)
1. Composite quasi-likelihood for high-dimensional variable selection | JRSSB, 73(3), 325-349, (2011)
Model selection robust and adaptive to the error distribution
(with Jianqing Fan and Weiwei Wang)