|
University of Bath |
Professor Christopher Jennison pursues research into the design and analysis of clinical trials. He has worked on sequential methods, which are used to monitor trials and make decisions on when to stop a study. His book with Professor Bruce Turnbull, "Group Sequential Methods with Applications to Clinical Trials", is a standard text on this topic, widely used by practicing statisticians. More recently, he has worked on adaptive clinical trial designs which allow a broader range of interim decision making, such as treatment selection or re-defining the target population, during the course of a trial. This research is informed by experience of clinical trial analysis at the Dana Farber Cancer Institute, Boston, consultancy with medical research institutes and pharmaceutical companies, and participation in clinical trial data monitoring committees. He has been a member of the DIRECT consortium (funded by the EU Innovative Medicines Initiative) and the MASTERMIND project (funded by the MRC and ABPI), which explored personalised treatment strategies for patients with type 2 diabetes. Professor Jennison is also interested in statistical image analysis, spatial statistics and the computational methods used to fit complex models to large data sets, for which the applications range from medical and biological data to remote sensing and modelling the atmosphere. |
|
|
|
|
American University |
Prof. Baron conducts research in sequential analysis, change-point problems, and Bayesian inference, occasionally applying results in epidemiology, clinical trials, insurance, energy finance, and semiconductor manufacturing. He is credited for introducing asymptotically pointwise optimal stopping rules in change-point detection, translating the corresponding classical concept of Bickel and Yahav from sequential estimation. The method was extended to dependent data, with nuisance parameters, multiple data streams, and almost arbitrary prior distributions. M. Baron elaborated several classes of sequentially planned statistical procedures. These are flexible group sequential sampling schemes with dynamically determined group sizes that result in substantial cost saving in cases where there is a fixed cost component, or the cost function is nonlinear. Group sequential clinical trials can be accelerated by this method, without sacrificing the probabilities of Type I and Type II errors. Baron participated in the design and analysis of several clinical trials. He authored a probability and statistics textbook for computer scientists and co-authored a series of books studying applications of statistics in sociology and marketing, classifying and exploring lifestyles and consumer behaviors. M. Baron is a Fellow of the American Statistical Association and a recipient of Abraham Wald award. |
|
|
|
|
Stanford University |
Tze Leung Lai is the Ray Lyman Wilbur Professor of Statistics and, by courtesy, of Biomedical Data Science and Computational & Mathematical Engineering at Stanford University, where he is also a faculty affiliate of Stanford’s new Doerr School of Sustainability & Climate Change and responsible for developing new courses on AI, machine learning, natural language processing, blockchains and cryptocurrencies. He was the inaugural recipient of the Abraham Wald Prize (in 2005) and has supervised over 77 PhD theses at Columbia and Stanford. He is well known for his work in adaptive design, risk analytics and management in finance and insurance, clinical trial designs, stochastic optimization and multi-armed bandits, biomarkers, longitudinal data analysis with survival endpoints, martingale regression and structural changes. He is the director of the Financial and Risk Modeling Institute (FARM) and Co-Director of the Center for Innovative Study Design (CISD) at Stanford’s School of Medicine (SoM). He is also an incoming Co-Director of Stanford’s Graduate School of Education. His talk will summarize his recent work, with former and current PhD students, on an efficient, automated MCMC (Markov Chain Monte Carlo) scheme for joint state and parameter estimation in hidden Markov models with multifaceted applications including robotics, deep learning, computerized education testing in education & mental health counseling. |
|
Georgia Tech |
Prof. Mei’s research focuses on statistics and machine learning, particularly, sequential analysis, change-point detection and streaming data analysis, and their applications to engineering, operation research, biomedical and health sciences. In his earlier career, Dr. Mei influenced the theoretical development on the foundation of sequential analysis and change-point detection, especially when the data could be dependent or not identically distributed. Later, motivated by sensor networks in engineering and biosurveillance, he has pioneered the development of updated theory of sequential analysis and change-point detection in the context of online monitoring high-dimensional data. He proposed the first family of scalable high-dimensional change-point detection schemes by raising a global alarm based on the sum of the local CUSUM statistics and also developing its asymptotic optimal properties. He also worked with his Ph.D. students to develop general scalable methodologies that apply shrinkage methods to address the sparsity issues when online monitoring high-dimensional data. Currently he is interested in developing useful theories and scalable methodologies for efficient real-time or online data-driven decision-making under various constraints on computing, communication, sampling, or privacy. Dr. Mei’s work has received several recognitions, including 2009 Abraham Wald prize in Sequential Analysis, 2010 NSF CAREER Award, and multiple best paper awards. |
|
Connecticut College |
Professor Zhuang obtained her doctoral degree in 2018 under the supervision of Professor Nitis Mukhopadhyay. She conducts research in theory and applications in statistics, including applied probability, sequential analysis, statistical inference, and interdisciplinary research at Connecticut College, New London, Connecticut. She has published more than 16 peer-reviewed full-length journal articles on sufficiency to ancillarity to distribution theory to sequential estimation and tests, along with deep collaborative research in leading international journals. She has developed statistical methodologies to make the statistical point and interval estimation and hypothesis testing more efficient and accurate under different scenarios. In applications of statistics, Professor Zhuang has completed a large array of interdisciplinary collaborative research projects with researchers from areas such as chemistry, environmental engineering, agriculture, and food science. She has organized and chaired 10 invited sessions at various international and regional conferences including the International Workshop in Sequential Methodologies (IWSM) and the New England Statistical Symposium (NESS). She has delivered more than 10 full-length invited research presentations at international and regional conferences as well as departmental colloquia. IWSM warmly welcomes this bright and young researcher to present a major lecture. |
|
Georgia Tech |
Dr. Yao Xie's research lies at the interface of statistics, optimization, and machine learning for data science problems. She worked on developing computationally efficient and statistically powerful algorithms with guarantees for engineering problems arising from various real-world applications through close interaction with domain experts. Her research interests are particularly in sequential data such as time series data, point process data, spatio-temporal data, and temporal dynamic networks, and in developing sequential prediction and change-point detection algorithms for such data. She has also worked on developing distribution-free high-dimensional hypothesis tests, conformal prediction, and uncertainty quantification for machine learning algorithms. Her work has generated societal and policy impacts, e.g., our data-driven police zone design was implemented by the Atlanta Police Department in 2019 and shown to be effective, which was selected as the INFORMS Wagner Prize Finalist in 2021. She received the National Science Foundation (NSF) CAREER Award in 2017, and the INFORMS Gaver Early Career Award for Excellence in Operations Research in 2022. She is currently an Associate Editor for IEEE Transactions on Information Theory, IEEE Transactions on Signal Processing, Sequential Analysis: Design Methods and Applications, INFORMS Journal on Data Science, and serves on the Editorial Board of the Journal of Machine Learning Research, and an Area Chair of NeurIPS in 2021 and 2022. |
