Floyd Demmon

PRACTITIONER: Experienced practitioner of wide variety of statistical methods from basic descriptive statistics through multiple linear and logistic regressions. Familiar with both parametric and non-parametric methods, and have applied basic data mining methods (classification and regression trees and, recently Bayesian Belief Networks) to various problems of automated inspection and surface defects from continuous cast steel and to production quality issues where processes have high degree of multi-collinearity. Have used the tools over the years to solve intractable problems and, above all, to reduce costs, either by eliminating non value-added work, or by improving product/process quality. Estimate total savings contributed over course of career to be in excess of over $4.5 million, and played key roles in other projects that saved at least twice as much in addition. TEACHER: Started teaching SPC, basic statistics, and designed experiments in the early 90's and continue to this day. Have developed and taught internal seminars on implementing SPC in automated (autocorrelated) processes, basic problem solving techniques for recent hires, and Minitab basics. Have coauthored/presented several papers at cokemaking conferences, at a continuous casting conference, at a DOE conference at Rochester Institute of Technology, at my local chapter of ASQ, and at the 4th Annual BayesiaLab User's Conference. Believe the reasons for learning are to benefit society and to share the knowledge with those coming after. LIFE LONG LEARNER: Current interests are in learning Bayesian statistical methods and, especially, Bayesian Networks to be able to attack the most complex issues facing my employer more effectively. Having never been especially good at math, the challenges will be many for a "late career" person, but I believe that it's when you stop trying to run that you begin to lose the ability to walk. Never stop learning.