Does anyone know how to plan a multi-level fractional factorial design of experiments?
I have 3 factors with 5 levels each, i.e.: 5³ = 125 experiments for a full factorial design, and I would like to reduce it to 45 experiments.
Matlab provides functions for multi-level full factorial and 2-level fractional factorial, but I don't know any function for multi-level fractional factorial designs...
As Mattias said. Try D-optimal design, but if you want to get specific combinations of your factors in design matrix then you should change the factors from continuous to discrete. This allows you to conduct your study at the desired factor combinations within the region of interest.
For those who might be interested:
A multi-level fractional factorial design of experiments using Matlab with 3 factors, 5 levels and 45 runs.
MLFF = cordexch(3,45,'linear','categorical',[1 2 3],'levels',[5 5 5])
I found this course in Coursera by Prof. Montgomery (Arizona State University).
The course is about Fractional Factorial Designs.
https://www.coursera.org/learn/factorial-fractional-factorial-designs?recoOrder=1&utm_medium=email&utm_source=recommendations&utm_campaign=khagQC9pEeuITlNXLNh0Mg
- Badges
- Science topic
- Similar topics
- Psychology