Skip to content Skip to navigation

Some Emerging Research Areas in Response Surface Methodology

Event Type: 
Seminar
Speaker: 
Christine Anderson-Cook
Event Date: 
Friday, September 19, 2014 -
12:00pm to 1:00pm
Location: 
SMLC 356
Audience: 
General PublicFaculty/StaffStudentsAlumni/Friends
Sponsor/s: 
stats group

Event Description: 

Response surface methodology is a collection of statistical
techniques for collecting and analyzing data through experimentation with
the goal of improving or optimizing processes and products. Since the first
book on this topic in the seventies (Myers, 1976) through today (Myers,
Montgomery and Anderson-Cook, 2009), the tools and methods developed have
been steadily increasing in number and becoming more sophisticated. In this
talk we explore several new research areas that have emerged in the past
few years. On the designed experiment side, these include: (1) Definitive
screening designs (Jones & Nachtsheim, 2011) represent a new class of
screening designs with some capability to estimate second order effects. By
blurring the lines between designs for first order models with those for
second order models, RSM has moved into a new era of considering designs
for multiple objectives. (2) Pareto front optimization provides a framework
for simultaneously considering competing objectives and finding a design
which balances several criterion bases on a weighting specified by the
experimenter. We consider both choosing a best design from a catalog of
designs and generating new designs using search algorithms subject to
experiment constraints.

On the analysis side, (3) finding optimal operating conditions based
multiple responses has also evolved to allow greater user flexibility in
specifying the relative prioritization of objectives and incorporating
uncertainty associated with estimating the responses. We present several
examples of these new research areas from the literature.

Event Contact

Contact Name: Yan Lu

Contact Phone: 5052772544

Contact Email: luyan@math.unm.edu