Xiaodong Yan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xiaodong Yan has a total of 45 co-authors. Here are the top co-authors by number of shared publications:
- Linglong Kong
University of Alberta
3 shared publications Bei Jiang
Department of Mathematical and Statistical Sciences, University of Alberta
2 shared publications- Jingxin Yan
Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing 100190, China
2 shared publications - Wei Wang
School of Statistics and Mathematics, Shandong University of Finance and Economics, Jinan 250014, China
2 shared publications Xianwen Ding
Department of Statistics Jiangsu University of Technology Changzhou 213001 Jiangsu People's Republic of China
2 shared publications- Anne-Sophie Charest
Université Laval
1 shared publication
Latest Publications
- Open Access
- 30 May 2024
- Published by MDPI AG in Mathematics
- Vol. 12 (11) , 1713
Demystifying the Two-Armed Futurity Bandit’s Unfairness and Apparent Fairness
While a gambler may occasionally win, continuous gambling inevitably results in a net loss to the casino. This study experimentally demonstrates the profitability of a particularly deceptive casino game: a two-armed antique Mills Futurity slot machine. The main findings clearly show that both non-random and random two-arm strategies, predetermined by the player and repeated without interruption, are always profitable for the casino, despite two coins being refunded for every two consecutive losses by the gambler. We theoretically explore the cyclical nature of slot machine strategies and speculate on the impact of the frequency of switching strategies on casino returns. Our results not only assist casino owners in developing and improving casino designs, but also guide gamblers to participate more cautiously in gambling.
- 30 May 2024
- Published by MDPI AG in Mathematics
- Vol. 12 (11) , 1713
Open AccessDemystifying the Two-Armed Futurity Bandit’s Unfairness and Apparent Fairness
While a gambler may occasionally win, continuous gambling inevitably results in a net loss to the casino. This study experimentally demonstrates the profitability of a particularly deceptive casino game: a two-armed antique Mills Futurity slot machine. The main findings clearly show that both non-random and random two-arm strategies, predetermined by the player and repeated without interruption, are always profitable for the casino, despite two coins being refunded for every two consecutive losses by the gambler. We theoretically explore the cyclical nature of slot machine strategies and speculate on the impact of the frequency of switching strategies on casino returns. Our results not only assist casino owners in developing and improving casino designs, but also guide gamblers to participate more cautiously in gambling.
- Open Access
Fast Fusion Clustering via Double Random Projection
In unsupervised learning, clustering is a common starting point for data processing. The convex or concave fusion clustering method is a novel approach that is more stable and accurate than traditional methods such as k-means and hierarchical clustering. However, the optimization algorithm used with this method can be slowed down significantly by the complexity of the fusion penalty, which increases the computational burden. This paper introduces a random projection ADMM algorithm based on the Bernoulli distribution and develops a double random projection ADMM method for high-dimensional fusion clustering. These new approaches significantly outperform the classical ADMM algorithm due to their ability to significantly increase computational speed by reducing complexity and improving clustering accuracy by using multiple random projections under a new evaluation criterion. We also demonstrate the convergence of our new algorithm and test its performance on both simulated and real data examples.
Open AccessFast Fusion Clustering via Double Random Projection
In unsupervised learning, clustering is a common starting point for data processing. The convex or concave fusion clustering method is a novel approach that is more stable and accurate than traditional methods such as k-means and hierarchical clustering. However, the optimization algorithm used with this method can be slowed down significantly by the complexity of the fusion penalty, which increases the computational burden. This paper introduces a random projection ADMM algorithm based on the Bernoulli distribution and develops a double random projection ADMM method for high-dimensional fusion clustering. These new approaches significantly outperform the classical ADMM algorithm due to their ability to significantly increase computational speed by reducing complexity and improving clustering accuracy by using multiple random projections under a new evaluation criterion. We also demonstrate the convergence of our new algorithm and test its performance on both simulated and real data examples.
- Open Access
- 24 March 2024
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 38 (19) , 21206-21213
Analysis of Differentially Private Synthetic Data: A Measurement Error Approach
Differentially private (DP) synthetic datasets have been receiving significant attention from academia, industry, and government. However, little is known about how to perform statistical inference using DP synthetic datasets. Naive approaches that do not take into account the induced uncertainty due to the DP mechanism will result in biased estimators and invalid inferences. In this paper, we present a class of maximum likelihood estimator (MLE)-based easy-to-implement bias-corrected DP estimators with valid asymptotic confidence intervals (CI) for parameters in regression settings, by establishing the connection between additive DP mechanisms and measurement error models. Our simulation shows that our estimator has comparable performance to the widely used sufficient statistic perturbation (SSP) algorithm in some scenarios but with the advantage of releasing a synthetic dataset and obtaining statistically valid asymptotic CIs, which can achieve better coverage when compared to the naive CIs obtained by ignoring the DP mechanism.
- 24 March 2024
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 38 (19) , 21206-21213
Open AccessAnalysis of Differentially Private Synthetic Data: A Measurement Error Approach
Differentially private (DP) synthetic datasets have been receiving significant attention from academia, industry, and government. However, little is known about how to perform statistical inference using DP synthetic datasets. Naive approaches that do not take into account the induced uncertainty due to the DP mechanism will result in biased estimators and invalid inferences. In this paper, we present a class of maximum likelihood estimator (MLE)-based easy-to-implement bias-corrected DP estimators with valid asymptotic confidence intervals (CI) for parameters in regression settings, by establishing the connection between additive DP mechanisms and measurement error models. Our simulation shows that our estimator has comparable performance to the widely used sufficient statistic perturbation (SSP) algorithm in some scenarios but with the advantage of releasing a synthetic dataset and obtaining statistically valid asymptotic CIs, which can achieve better coverage when compared to the naive CIs obtained by ignoring the DP mechanism.
- Open Access
- 8 August 2023
- Published by Wiley in Canadian Journal of Statistics
- Vol. 52 (2) , 618-635
High‐dimensional model averaging for quantile regression
- 8 August 2023
- Published by Wiley in Canadian Journal of Statistics
- Vol. 52 (2) , 618-635
Open AccessHigh‐dimensional model averaging for quantile regression
- 13 July 2023
- Published by Elsevier in Economics Letters
- Vol. 230
A bi-integrative analysis of two-dimensional heterogeneous panel data models
- 13 July 2023
- Published by Elsevier in Economics Letters
- Vol. 230
A bi-integrative analysis of two-dimensional heterogeneous panel data models
- Open Access
- 26 June 2023
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 37 (6) , 7270-7278
Opposite Online Learning via Sequentially Integrated Stochastic Gradient Descent Estimators
Stochastic gradient descent algorithm (SGD) has been popular in various fields of artificial intelligence as well as a prototype of online learning algorithms. This article proposes a novel and general framework of one-sided testing for streaming data based on SGD, which determines whether the unknown parameter is greater than a certain positive constant. We construct the online-updated test statistic sequentially by integrating the selected batch-specific estimator or its opposite, which is referred to opposite online learning. The batch-specific online estimators are chosen strategically according to the proposed sequential tactics designed by two-armed bandit process. Theoretical results prove the advantage of the strategy ensuring the distribution of test statistic to be optimal under the null hypothesis and also supply the theoretical evidence of power enhancement compared with classical test statistic. In application, the proposed method is appealing for statistical inference of one-sided testing because it is scalable for any model. Finally, the superior finite-sample performance is evaluated by simulation studies.
- 26 June 2023
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 37 (6) , 7270-7278
Open AccessOpposite Online Learning via Sequentially Integrated Stochastic Gradient Descent Estimators
Stochastic gradient descent algorithm (SGD) has been popular in various fields of artificial intelligence as well as a prototype of online learning algorithms. This article proposes a novel and general framework of one-sided testing for streaming data based on SGD, which determines whether the unknown parameter is greater than a certain positive constant. We construct the online-updated test statistic sequentially by integrating the selected batch-specific estimator or its opposite, which is referred to opposite online learning. The batch-specific online estimators are chosen strategically according to the proposed sequential tactics designed by two-armed bandit process. Theoretical results prove the advantage of the strategy ensuring the distribution of test statistic to be optimal under the null hypothesis and also supply the theoretical evidence of power enhancement compared with classical test statistic. In application, the proposed method is appealing for statistical inference of one-sided testing because it is scalable for any model. Finally, the superior finite-sample performance is evaluated by simulation studies.
- Open Access
- 26 June 2023
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 37 (8) , 10104-10112
Spearman Rank Correlation Screening for Ultrahigh-Dimensional Censored Data
Herein, we propose a Spearman rank correlation-based screening procedure for ultrahigh-dimensional data with censored response cases. The proposed method is model-free without specifying any regression forms of predictors or response variables and is robust under the unknown monotone transformations of these response variable and predictors. The sure-screening and rank-consistency properties are established under some mild regularity conditions. Simulation studies demonstrate that the new screening method performs well in the presence of a heavy-tailed distribution, strongly dependent predictors or outliers, and offers superior performance over the existing nonparametric screening procedures. In particular, the new screening method still works well when a response variable is observed under a high censoring rate. An illustrative example is provided.
- 26 June 2023
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 37 (8) , 10104-10112
Open AccessSpearman Rank Correlation Screening for Ultrahigh-Dimensional Censored Data
Herein, we propose a Spearman rank correlation-based screening procedure for ultrahigh-dimensional data with censored response cases. The proposed method is model-free without specifying any regression forms of predictors or response variables and is robust under the unknown monotone transformations of these response variable and predictors. The sure-screening and rank-consistency properties are established under some mild regularity conditions. Simulation studies demonstrate that the new screening method performs well in the presence of a heavy-tailed distribution, strongly dependent predictors or outliers, and offers superior performance over the existing nonparametric screening procedures. In particular, the new screening method still works well when a response variable is observed under a high censoring rate. An illustrative example is provided.
- 14 June 2023
- Published by Oxford University Press (OUP) in Journal of the Royal Statistical Society Series B: Statistical Methodology
- Vol. 85 (4) , 1271-1298
Strategic two-sample test via the two-armed bandit process
This study aims to improve the power of two-sample tests by analysing whether the difference between two population parameters is larger than a prespecified positive equivalence margin. The classic test statistic treats the original data as exchangeable, while the proposed test statistic breaks the structure and proposes employing a two-armed bandit process to strategically integrate the data and thus a strategy-specific test statistic is constructed by combining the classic CLT with the law of large numbers. The developed asymptotic theory is investigated by using nonlinear limit theory in a larger probability space and relates to the ‘strategic CLT’ with a clearly defined density function. The asymptotic distribution demonstrates that the proposed statistic is more concentrated under the null hypothesis and less concentrated under the alternative than the classic CLT, thereby enhancing the testing power. Simulation studies provide supporting evidence for the theoretical results and portray a more powerful performance when using finite samples. A real example is also added for illustration.
- 14 June 2023
- Published by Oxford University Press (OUP) in Journal of the Royal Statistical Society Series B: Statistical Methodology
- Vol. 85 (4) , 1271-1298
Strategic two-sample test via the two-armed bandit process
This study aims to improve the power of two-sample tests by analysing whether the difference between two population parameters is larger than a prespecified positive equivalence margin. The classic test statistic treats the original data as exchangeable, while the proposed test statistic breaks the structure and proposes employing a two-armed bandit process to strategically integrate the data and thus a strategy-specific test statistic is constructed by combining the classic CLT with the law of large numbers. The developed asymptotic theory is investigated by using nonlinear limit theory in a larger probability space and relates to the ‘strategic CLT’ with a clearly defined density function. The asymptotic distribution demonstrates that the proposed statistic is more concentrated under the null hypothesis and less concentrated under the alternative than the classic CLT, thereby enhancing the testing power. Simulation studies provide supporting evidence for the theoretical results and portray a more powerful performance when using finite samples. A real example is also added for illustration.
- Open Access
- 15 February 2023
- Published by Springer Nature in BMC Medical Research Methodology
- Vol. 23 (1) , 1-13
Propensity score analysis with missing data using a multi-task neural network
Propensity score analysis is increasingly used to control for confounding factors in observational studies. Unfortunately, unavoidable missing values make estimating propensity scores extremely challenging. We propose a new method for estimating propensity scores in data with missing values. Both simulated and real-world datasets are used in our experiments. The simulated datasets were constructed under 2 scenarios, the presence (T = 1) and the absence (T = 0) of the true effect. The real-world dataset comes from LaLonde’s employment training program. We construct missing data with varying degrees of missing rates under three missing mechanisms: MAR, MCAR, and MNAR. Then we compare MTNN with 2 other traditional methods in different scenarios. The experiments in each scenario were repeated 20,000 times. Our code is publicly available at https://github.com/ljwa2323/MTNN. Under the three missing mechanisms of MAR, MCAR and MNAR, the RMSE between the effect and the true effect estimated by our proposed method is the smallest in simulations and in real-world data. Furthermore, the standard deviation of the effect estimated by our method is the smallest. In situations where the missing rate is low, the estimation of our method is more accurate. MTNN can perform propensity score estimation and missing value filling at the same time through shared hidden layers and joint learning, which solves the dilemma of traditional methods and is very suitable for estimating true effects in samples with missing values. The method is expected to be broadly generalized and applied to real-world observational studies.
- 15 February 2023
- Published by Springer Nature in BMC Medical Research Methodology
- Vol. 23 (1) , 1-13
Open AccessPropensity score analysis with missing data using a multi-task neural network
Propensity score analysis is increasingly used to control for confounding factors in observational studies. Unfortunately, unavoidable missing values make estimating propensity scores extremely challenging. We propose a new method for estimating propensity scores in data with missing values. Both simulated and real-world datasets are used in our experiments. The simulated datasets were constructed under 2 scenarios, the presence (T = 1) and the absence (T = 0) of the true effect. The real-world dataset comes from LaLonde’s employment training program. We construct missing data with varying degrees of missing rates under three missing mechanisms: MAR, MCAR, and MNAR. Then we compare MTNN with 2 other traditional methods in different scenarios. The experiments in each scenario were repeated 20,000 times. Our code is publicly available at https://github.com/ljwa2323/MTNN. Under the three missing mechanisms of MAR, MCAR and MNAR, the RMSE between the effect and the true effect estimated by our proposed method is the smallest in simulations and in real-world data. Furthermore, the standard deviation of the effect estimated by our method is the smallest. In situations where the missing rate is low, the estimation of our method is more accurate. MTNN can perform propensity score estimation and missing value filling at the same time through shared hidden layers and joint learning, which solves the dilemma of traditional methods and is very suitable for estimating true effects in samples with missing values. The method is expected to be broadly generalized and applied to real-world observational studies.
- 28 December 2022
- Published by Taylor & Francis in Statistics
- Vol. 57 (1) , 26-52
Model averaging for generalized linear models with missing at random covariates
We consider a frequentist model averaging approach for prediction issue of generalized linear model with covariates involving missing at random. To avoid the uncertainty problem from imputation on the missing covariates data, this work adjusts model averaging criterion by selecting a batch of candidate models with the available covariate patterns rather than all possible combinations of covariates. The weights over the batch of candidate models are estimated based on the leave-one-out cross-validation criterion of the fully observed data. The asymptotic optimality of the proposed method is established under some regular conditions. Simulation studies are carried out to demonstrate the finite sample performance of our proposed approach in comparison to other methods. A real data set from NHANES 2017–2018 is also applied to check the effectiveness of the proposed approach.
- 28 December 2022
- Published by Taylor & Francis in Statistics
- Vol. 57 (1) , 26-52
Model averaging for generalized linear models with missing at random covariates
We consider a frequentist model averaging approach for prediction issue of generalized linear model with covariates involving missing at random. To avoid the uncertainty problem from imputation on the missing covariates data, this work adjusts model averaging criterion by selecting a batch of candidate models with the available covariate patterns rather than all possible combinations of covariates. The weights over the batch of candidate models are estimated based on the leave-one-out cross-validation criterion of the fully observed data. The asymptotic optimality of the proposed method is established under some regular conditions. Simulation studies are carried out to demonstrate the finite sample performance of our proposed approach in comparison to other methods. A real data set from NHANES 2017–2018 is also applied to check the effectiveness of the proposed approach.