FTC 2023 Short Courses

Short courses will be held on October 4, 2023. Attendees may registers for the conference and short courses together or separately. Conference attendance is not required to attend a short course.

Full Day

Peter Goos – OMARS Designs: Modern Experimental Designs for Simultaneous Screening and Response Surface Optimization

Yufeng Liu – Introduction to Statistical Machine Learning using R

James Lucas – Advanced Quality Control Methods: CUSUM and EWMA Procedures

Half Day

Robert B. Gramacy (Bobby) – Gaussian Process Modeling, Design and Optimization for the Applied Sciences


8:30am to 5:30pm (1 hour lunch at 12:30pm) 

OMARS Designs: Modern Experimental Designs for Simultaneous Screening and Response Surface Optimization

Peter Goos, KU Leuven and University of Antwerp

Abstract: Traditionally, factor screening and response surface experimentation have always been performed in two separate steps. The increasing pressure to speed up innovation necessitates an adaptation of that strategy, by combining factor screening and optimization into a single experiment. In recent years, suitable economical experimental designs have been developed to study many factors simultaneously. These designs are called Orthogonal Minimally Aliased Response Surface or OMARS designs. The family of OMARS designs forms a major extension of the family of definitive screening designs. Due to the fact that they involve three levels per factor, their larger power for detecting quadratic effects and their good projection properties, OMARS designs serve as screening experiment and as response surface experiments at the same time.

In this short course, we first discuss the properties of OMARS designs, as well as various design quality characteristics that can be used to select OMARS designs that are appropriate for a given problem. A key aspect in our design selection is the use of multiple criteria to select a good experimental design, rather than a single design optimality criterion. We also discuss extensions of the family of OMARS designs. For instance, we introduce mixed-level OMARS designs which allow the study of two-level factors in addition to three-level factors. We also explain how OMARS designs can be arranged in blocks, so that the resulting designs are orthogonally blocked. Finally, we introduce the concept of strong OMARS designs, which are economical orthogonal designs with the same orthogonality properties as central composite and Box-Behnken designs, but with fewer runs and a higher information content. Throughout the course, we will demonstrate a brand new web-based application to explore the catalog of OMARS designs.

Target audience and prerequisites: This course is for anyone interested in conducting cost-efficient experiments in business and industry.  Prior familiarity with classical screening experiments, definitive screening designs and response surface designs will be helpful.

CONTACT: peter.goos@kuleuven.be

Peter Goos is a full professor at the Faculty of Bio-Science Engineering of KU Leuven, and at the Faculty of Business and Economics of the University of Antwerp, where he teaches various introductory and advanced courses on statistics and probability. His main research area is the statistical design and analysis of experiments. Besides numerous influential articles in various kinds of scientific journals, he has published the books The Optimal Design of Blocked and Split-Plot Experiments, Optimal Experimental Design: A Case-Study Approach, Statistics with JMP: Graphs, Descriptive Statistics and Probability and Statistics with JMP: Hypothesis Tests, ANOVA and Regression. For his work, Peter Goos has received three Shewell Awards, two Lloyd S. Nelson Awards and a Brumbaugh Award from the American Society for Quality, the Ziegel Award and the Statistics in Chemistry Award from the American Statistical Association, and the Young Statistician Award of the European Network for Business and Industrial Statistics (ENBIS). Peter is known for this ability to introduce new design of experiments concepts in an accessible fashion to non-academics.

Introduction to Statistical Machine Learning using R

Yufeng Liu, University of North Carolina

Abstract: Statistical machine learning is an interdisciplinary research area which is closely related to statistics, computer sciences, engineering, and bioinformatics. Many statistical machine learning techniques and algorithms have proven to be very useful for various scientific areas. This course will provide an overview of statistical machine learning and data mining techniques with applications to the analysis of real-world data. Supervised learning techniques include penalized regression such as LASSO and its variants, support vector machines, and tree-based methods. Unsupervised learning techniques include dimension reduction methods such as principal components analysis, and clustering analysis. The main emphasis will be on the analysis of real data sets from various scientific fields. The techniques discussed will be demonstrated in R.

Target audience and prerequisites: This course is intended for researchers who have some knowledge of statistics and want to be introduced to statistical machine learning and data mining, or practitioners who would like to apply statistical machine learning techniques to their problems. Participants should be familiar with linear regression and basic statistical and probability concepts, as well as some familiarity with R programming.

CONTACT: yfliu@email.unc.edu

Yufeng Liu is currently professor in Department of Statistics and Operations Research, Department of Biostatistics, and Department of Genetics at UNC-Chapel Hill. His current research interests include statistical machine learning, high dimensional data analysis, personalized medicine, and bioinformatics. He has taught statistical machine learning courses multiple times at UNC, as well as short courses on this subject at Joint Statistical Meetings, ENAR, FDA, and Biostatistics Summer Institutes at University of Washington. Dr. Liu serves on the editorial boards for several statistics journals, including as Area Editor for Annals of Applied Statistics, and Associate Editor for Journal of American Statistical Association. He received the CAREER Award from National Science Foundation in 2008, and Ruth and Phillip Hettleman Prize for Artistic and Scholarly Achievement in 2010, and the inaugural Leo Breiman Junior Award in 2017. Dr. Liu is currently an elected fellow at American Statistical Association, Institute of Mathematical Statistics (IMS), and an elected member of International Statistical Institute.

Advanced Quality Control Methods: CUSUM and EWMA Procedures

James Lucas, J. Lucas & Associates

This course will discuss CUSUM (Cumulative Summation) and EWMA (Exponentially Weighted Moving Average) Procedures for Quality Control Applications. The presenter is a consultant and researcher who was an integral part of the largest known implementation of CUSUM procedures. He has guided many successful implementations of these procedures and was the developer of many of the most used enhancements. This will be a “hands on” course with discussions of real problems and workshops to practice using the techniques. You will come from this course knowing how to use these procedures to monitor a process (and to detect problems). The course knowledge will enhance your data analysis skills. Both variables and count data (including procedures for rare events) will be discussed. Comparisons with classical Shewhart procedures will be given throughout the course to show where these advanced procedures have large benefits. The prerequisite for the course is a quality control course or some implementation experience.

Target audience and prerequisites:

CONTACT: JamesM.Lucas@verizon.net

James Lucas


Gaussian Process Modeling, Design and Optimization for the Applied Sciences
1:30pm to 5:30pm

Robert B. Gramacy (Bobby), Virginia Tech

Abstract: This course details statistical techniques at the interface between geostatistics, machine learning, mathematical modeling via computer simulation, calibration of computer models to data from field experiments, and model-based sequential design and optimization under uncertainty (a.k.a. Bayesian Optimization). The treatment will include some of the historical methodology in the literature, and canonical examples, but will primarily concentrate on modern statistical methods, computation and implementation, as well as modern application/data type and size. The course will return at several junctures to real-word experiments coming from the physical, biological and engineering sciences, such as studying the aeronautical dynamics of a rocket booster re-entering the atmosphere; modeling the drag on satellites in orbit; designing a hydrological remediation scheme for water sources threatened by underground contaminants; studying the formation of super-nova via radiative shock hydrodynamics; modeling the evolution a spreading epidemic. The course material will emphasize deriving and implementing methods over proving theoretical properties.

Target audience and prerequisites: The target audience is research professionals wishing to learn about modern techniques in process improvement, computer simulation experiments, and the synthesis of virtual and observational data. Students will benefit from exposure to computational linear algebra, design of experiments, ordinary linear regression, and familiarity with a programming language (e.g, R, Matlab, Python). None of these, however, will be pivotal to extracting value from the course. The presentation will be in Rmarkdown, which means that all slides and analyses will be reproducible, and be accompanied by a full scale, and fully open source textbook, also authored in Rmarkdown.

Free Electronic Textbook: https://bobby.gramacy.com/surrogates/

CONTACT: rbg@vt.edu

Dr. Robert B. Gramacy is a Professor of Statistics in the College of Science at Virginia Polytechnic and State University (Virginia Tech). Previously he was an Associate Professor of Econometrics and Statistics at the Booth School of Business, and a fellow of the Computation Institute at The University of Chicago. His research interests include Bayesian modeling methodology, statistical computing, Monte Carlo inference, nonparametric regression, sequential design, and optimization under uncertainty. He currently serves as the Editor-elect at Technometrics, an ASA journal, and as President for the ASA’s Uncertainty Quantification Interest Group. Recently he completed tours as President of the ASA’s Section on Physical and Engineering Sciences, and as Treasurer for the International Society of Bayesian Analysis

Professor Gramacy is a computational statistician. He specializes in areas of real-data analysis where the ideal modeling apparatus is impractical, or where the current solutions are inefficient and thus skimp on fidelity. Such endeavors often require new models, new methods, and new algorithms. His goal is to be impactful in all three areas while remaining grounded in the needs of a motivating application. His aim is to release general purpose software for consumption by the scientific community at large, not only other statisticians. He is the primary author on six R packages available on CRAN, two of which (tgp, and monomvn) have won awards from statistical and practitioner communities.