A5088044292- Statistical Methods and Inference
- Advanced Statistical Methods and Models
- Advanced Bandit Algorithms Research
- Machine Learning and Algorithms
- Advanced Causal Inference Techniques
- Statistical Methods in Clinical Trials
- Stochastic Gradient Optimization Techniques
- Banana Cultivation and Research
- Sparse and Compressive Sensing Techniques
- Bioinformatics and Genomic Networks
- Plant-Microbe Interactions and Immunity
- Spectroscopy and Chemometric Analyses
- Optimization and Search Problems
- Gene Regulatory Network Analysis
- Renal cell carcinoma treatment
- Smart Grid Energy Management
- Bladder and Urothelial Cancer Treatments
- Educational Innovations and Challenges
- Advanced Algorithms and Applications
- Climate Change Policy and Economics
- Machine Learning in Healthcare
- Industrial Vision Systems and Defect Detection
- Metaheuristic Optimization Algorithms Research
- Digital Mental Health Interventions
- Advanced Fluorescence Microscopy Techniques
University of Wisconsin–Madison
2023-2024
Harvard University
2024
Princeton University
2019-2022
Fuzhou University
2020
State Key Laboratory For Conservation and Utilization of Subtropical Agro-Bioresources
2020
When the data are stored in a distributed manner, direct applications of traditional statistical inference procedures often prohibitive due to communication costs and privacy concerns. This paper develops investigates two Communication-Efficient Accurate Statistical Estimators (CEASE), implemented through iterative algorithms for optimization. In each iteration, node machines carry out computation parallel communicate with central processor, which then broadcasts aggregated information new...
In this article, we study the contextual dynamic pricing problem where market value of a product is linear in its observed features plus some noise. Products are sold one at time, and only binary response indicating success or failure sale observed. Our model setting similar to work by? except that expand demand curve semiparametric learn dynamically both parametric nonparametric components. We propose statistical learning decision making policy minimizes regret (maximizes revenue) by...
The escalating prevalence of cannabis use, and associated cannabis-use disorder (CUD), poses a significant public health challenge globally. With notably wide treatment gap, especially among emerging adults (EAs; ages 18-25), addressing use CUD remains pivotal objective within the 2030 United Nations Agenda for Sustainable Development Goals (SDG). In this work, we develop an online reinforcement learning (RL) algorithm called reBandit which will be utilized in mobile study to deliver...
We consider the problem of variance reduction in randomized controlled trials, through use covariates correlated with outcome but independent treatment. propose a machine learning regression-adjusted treatment effect estimator, which we call MLRATE. MLRATE uses predictors to reduce estimator variance. It employs cross-fitting avoid overfitting biases, and prove consistency asymptotic normality under general conditions. is robust poor predictions from step: if are uncorrelated outcomes,...
When the data are stored in a distributed manner, direct application of traditional statistical inference procedures is often prohibitive due to communication cost and privacy concerns. This paper develops investigates two Communication-Efficient Accurate Statistical Estimators (CEASE), implemented through iterative algorithms for optimization. In each iteration, node machines carry out computation parallel communicate with central processor, which then broadcasts aggregated information new...
The escalating prevalence of cannabis use, and associated cannabis-use disorder (CUD), poses a significant public health challenge globally. With notably wide treatment gap, especially among emerging adults (EAs; ages 18-25), addressing use CUD remains pivotal objective within the 2030 United Nations Agenda for Sustainable Development Goals (SDG). In this work, we develop an online reinforcement learning (RL) algorithm called reBandit which will be utilized in mobile study to deliver...
The escalating prevalence of cannabis use poses a significant public health challenge globally. In the U.S., is more prevalent among emerging adults (EAs) (ages 18-25) than any other age group, with legalization in multiple states contributing to perception that less risky prior decades. To address this growing concern, we developed MiWaves, reinforcement learning (RL) algorithm designed optimize delivery personalized intervention prompts reduce EAs. MiWaves leverages domain expertise and...
Online AI decision-making algorithms are increasingly used by digital interventions to dynamically personalize treatment individuals. These determine, in real-time, the delivery of based on accruing data. The objective this paper is provide guidelines for enabling effective monitoring online with goal (1) safeguarding individuals and (2) ensuring data quality. We elucidate discuss our experience two intervention clinical trials (Oralytics MiWaves). Our include developing fallback methods,...
We consider the contextual bandit problem where at each time, agent only has access to a noisy version of context and error variance (or an estimator this variance). This setting is motivated by wide range applications true for decision-making unobserved, prediction potentially complex machine learning algorithm available. When non-vanishing, classical algorithms fail achieve sublinear regret. propose first online in with regret guarantees under mild conditions. The key idea extend...
App-based N-of-1 trials offer a scalable experimental design for assessing the effects of health interventions at an individual level. Their practical success depends on strong motivation participants, which, in turn, translates into high adherence and reduced loss to follow-up. One way maintain participant engagement is by sharing their interim results. Continuously testing hypotheses during trial, known as "peeking", can also lead shorter, lower-risk detecting early. Nevertheless,...
本文将通过布尔网络模型研究这个问题.该模型已被成功运用于研究大型基因表达网络中一般性 或是全局性的性质 [17-24] .对细胞状态布尔形式的描述基于生物网络各组成部分所具有的 "开关" 特 征. 这种定性模型的优势在于, 它所做的预测对未知的动力学参数并不敏感 [25] .第 2 节将引入基本 概念, 并介绍传统布尔网络及其随机形式的拓展.第 3 节将介绍指数扰动马氏链的数学理论及其推广.第 4 和 5 节将这些理论应用于两个实际算例, 包括了一个人造算例和一个含有两个极限环的真实生 物网络 (节点: p53, Siah-1, β-catenin, p14/19 ARF, Mdm-2). 确定性与随机性布尔网络模型首先介绍一个应用广泛的布尔网络模型 [22] .给定正整数 N , S =
High-dimensional linear regression has been intensively studied in the community of statistics last two decades. For convenience theoretical analyses, classical methods usually assume independent observations and sub-Gaussian-tailed errors. However, neither them hold many real high-dimensional time-series data. Recently [Sun, Zhou, Fan, 2019, J. Amer. Stat. Assoc., press] proposed Adaptive Huber Regression (AHR) to address issue heavy-tailed They discover that robustification parameter loss...
Abstract Background: Banana Fusarium wilt is a devastating disease of bananas caused by oxysporum f. sp. cubense (Foc) and serious threat to the global banana industry. Knowledge pathogenic molecular mechanism interaction between host Foc limited. Results: In this study, we confirmed changes gene expression pathways in Cavendish variety ‘Brazilian’ during early infection with Foc1 Foc4 comparative transcriptomics analysis. 1862 226 differentially expressed genes (DEGs) were identified roots...
Multiple dynamic pathways always exist in biological networks, but their robustness against internal fluctuations and relative stability have not been well recognized carefully analyzed yet. Here we try to address these issues through an illustrative example, namely the Siah-1/beta-catenin/p14/19 ARF loop of protein p53 dynamics. Its deterministic Boolean network model predicts that two parallel with comparable magnitudes attractive basins should after is activated when a cell becomes...
An improved Otsu algorithm is proposed for corrosion detection around aircraft rivets. On the basis of traditional algorithm, which easy to cause loss some areas, maximum between-class variance value adjusted by adding weight k, and corresponding threshold obtained from relationship between value, so as realize optimal segmentation areas. The effect compared with mean square error, according condition, degree introduced classify grade. experimental results show that simple extract features...
Best subset selection (BSS) is widely known as the holy grail for high-dimensional variable selection. Nevertheless, notorious NP-hardness of BSS substantially restricts its practical application and also discourages theoretical development to some extent, particularly in current era big data. In this paper, we investigate properties when target sparsity greater than or equal true sparsity. Our main message that robust against design dependence terms achieving model consistency sure...
Abstract Background: Banana Fusarium wilt is a devastating disease of bananas caused by oxysporum f. sp. cubense (Foc) and serious threat to the global banana industry. Knowledge pathogenic molecular mechanism interaction between host Foc limited. Results: In this study, we confirmed changes gene expression pathways in Cavendish variety ‘Brazilian’ during early infection with Foc1 Foc4 comparative transcriptomics analysis. 1862 226 differentially expressed genes (DEGs) were identified roots...