Thursday, November 18 (Times listed in US Eastern Time)
9:00 – 10:00am Opening Remarks (Stefan Steiner) and Overview of Statistical Engineering Handbook (Lynne Hare)
Lynne Hare summarizes key features of the newly published first edition of the Statistical Engineering Handbook, available to ISEA members by logging in to the website: www.isea-change.org (Membership is free.) Discover thinking, enabling technologies, strategies and tactics, together with necessary quantitative and qualitative skills needed to efficiently garner actionable information from data sets, all directed to organizational improvement, however measured.
10:00 – 11:00am Keynote: Engineering Systems that Learn: The Need for Blends of Computation, Economics, and Statistics
by Michael I. Jordan (University of California, Berkeley)
Real-world applications of statistical inference increasingly take the form of large-scale systems that continually learn from data and interact with human decision-making. The design of such systems requires conjoined thinking that is simultaneously computational, economic, and statistical. I will discuss various research vignettes that bring these three thinking styles together, with particular emphasis on decision-making under uncertainty in the context of multiple decision-makers.
I will emphasize the major challenges that need to be faced in order for a field of "statistical engineering" to emerge and take its place along previous engineering disciplines such as civil, electrical, mechanical, and chemical engineering.
11:10 – 12:40pm The Past, Present, and Future of Statistical Engineering: A Panel Discussion
In 2012, Quality Engineering published a special issue devoted Statistical Engineering (co-edited by Christine Anderson-Cook and Lu Lu) that included two discussion papers in which invited panelists broadly discussed the past, present, and future of Statistical Engineering. Almost ten years later, we revisit this topic and convene a group of panelists to discuss whether the discipline has come since the 2012 special issue, including successes, challenges, and where the discipline goes from here. This session will feature the following moderator and panelists:
· Roger Hoerl, Union College (moderator)
· Christine Anderson-Cook, LANL
· Stefan Steiner, University of Waterloo
· Richard De Veaux, Williams College
· Diego Kuonen, Statoo Consulting & University of Geneva
· Allison Jones-Farmer, Miami University
· Jeroen de Mast, Jheronimus Academy of Data Science
· Michael Jordan, University of California, Berkeley
1:00 – 2:00 pm Roundtable Discussion – topics coming soon
Friday, November 19 (Times listed in US Eastern Time)
9:00 – 10:00am Solving Complex Problems with Statistical Engineering by Willis Jensen, W.L. Gore & Associates
At the inaugural 2018 Statistical Engineering Summit, I shared an example of a problem of determining Associate contribution at Gore using networks. This problem was solved by a dedicated team and illustrates concepts of statistical engineering. Since that time, we have continued to learn as we execute the deployed solution each year. I will revisit this problem and its solution which serves as an example of a complex problem. I’ll share a framework that describes complex problems in general and then illustrate how a Statistical Engineering approach can help solve this and other complex problems.
10:15 – 11:15am Guiding Complex Design of Experiment Choices to Match Requirements for Multiple Responses by Christine M. Anderson-Cook, Los Alamos National Laboratory
In a recent collaboration, material scientists requested assistance from statisticians to generate suitable designed experiments for studying multiple properties of an expensive synthetic material. Collecting data for each of the properties required different amounts of time and had vastly different associated costs. The analysis of each property was to be integrated into separate computer models and hence required different precision. This case study describes the process used by a team of scientists and statisticians to reach consensus on which data should be collected for each of the material properties conditional on available resources and material. Providing the right graphical summaries to highlight the trade-offs between alternative designs was key to building a common background of understanding and guiding productive discussions.
11:30 – 12:30 NASA’s Statistical Engineering Journey by Peter Parker, NASA
Statistical engineering within NASA first gained recognition by demonstrating significant technical impact, and those early successes initiated a journey where the discipline went from unknown, to rejected, to accepted, and finally sought after. Currently, the state of the discipline within NASA is stable and growing, however, reaching an aspired destination of infusing statistical engineering routinely in aerospace research and development to accelerate learning, ensure strategic resource investment, and inform decision making still lies ahead. This presentation offers a retrospective view of NASA’s journey of introducing and implementing statistical engineering in aeronautics, space exploration, and atmospheric science. It highlights milestones in technical and organizational impact, development of targeted educational resources, and methodological extensions that have grown its acceptance. The goal of the presentation is to share experience-based insights on NASA’s journey of leveraging the discipline of statistical engineering.
1:00 – 2:00 pm Roundtable Discussion – topics coming soon
Regular Registration Fee: $100
Academic Student Registration Fee: $50 – use promo code ACADEMIC50
(This applies to an individual who is currently engaged in full-time undergraduate or graduate level studies at a postsecondary institution.)