About our group

Nofima has a strong tradition within multivariate data analysis, statistics, sensometrics and chemometrics. Core members such as Tormod Næs and Harald Martens have contributed to the field of chemometrics since the 70’ties, in particular to partial least squares methodology (PLS) and multivariate calibration in general. Traditionally our applications have been centered on food science, sensometrics, process modeling and spectroscopy. Today we also work with applications within the –omics area (proteomics, metabolomics, functional genomics, microbiota) as well as in microbiology and health related research.

Currently our group comprises about 10 scientists with different backgrounds, but with common interests in multivariate data analysis.

 

We develop methods for use in

Data modelling at Nofima

Nofima has a strong tradition within multivariate data analysis, statistics, sensometrics and chemometrics. Core members such as Tormod Næs and Harald Martens have contributed to the field of chemometrics since the 70’ties, in particular to partial least squares methodology (PLS) and multivariate calibration in general. Traditionally our applications have been centered on food science, sensometrics, process modeling and spectroscopy. Today we also work with applications within the –omics area (proteomics, metabolomics, functional genomics, microbiota) as well as in microbiology and health related research.

Currently our group comprises about 10 scientists with different backgrounds, but with common interests in multivariate data analysis.

Multiblock data analysis

Today the major challenge within data analysis lies within handling increasing amounts of data. Frequently the data come from different measurement instruments (including spectroscopic measurements, sensory panel, consumer surveys etc.), are taken at different time points, at different biological levels or at different points in a process. These situations generate what we call multiblock data, i.e. there are several blocks of data for the same set of samples. We continue our tradition within multivariate analysis by developing and extending existing tools for analyzing multiblock data.

We have strong focus on the following topics:

  1. Exploratory data analysis
  2. Regression and classification
  3. Multivariate ANOVA
  4. Common and distinct components in multi-block analysis
  5. Multi-block regression and classification
  6. Variable selection (regression and classification)
  7. Categorical data
  8. Individual differences
  9. Unification of different approaches and relation between the methods