Sensometrics is the name for applying and developing statistics for the
analysis of human perceptual measurements. These are used in food and
many other industries for product development.
This is a general announcement of the possibility to do an interesting
and industrially relevant project within this area with
Per Bruun Brockhoff (PerBB) as supervisor.
PerBB developed methods and software during 25 years:
- Many projects with national and international companies
- Sharing open source softwares online (and training podcasts)
- R-packages sensR, ordinal, lmerTest, SensMixed, mumm
Possible project elements: (not all in each project)
- Danish food companies, Danish Meat Research Institute, Bang & Olufsen
- Unilever, UK, maybe other companies.
- Working with data from one or more companies
- “Proof of concepts” analysis on company data using DTU developed software and methods
- Further development of R-packages
- Learning about sensometrics
- Learning about and working with binomial data from discrimination and similarity testing (inkl. link to generalized linear models and mixed models) (R-package sensR)
- Ordinal regression models, basic and extended versions, mixed effect versions (R-package ordinal)
- Linear mixed models and extensions, (R-packages lmerTest, SensMixed and mumm)
- Chemometrics and multivariate regression methods such as PCA, PCR, PLS and the like. (In Sensory also called “preference mapping”)
In collaboration withDanish or international company
Some courses in statistics