Aim and content

In industry and research huge amounts of physical, chemical, sensory and other quality measurements are produced on all sorts of materials, processes and products. Exploratory data analysis / chemometrics offers a tool for extracting the optimal information from these data sets through the use of digitalization (modern software and computer technology).

The course will give a step-by-step theoretical introduction to exploratory data analysis / chemometrics supported by practical examples from food science, environmental science, pharmaceutical science etc.

Methods for exploratory analysis (Principal Component Analysis), multivariate calibration (Partial Least Squares) and basic data preprocessing are considered. The mathematics behind most of the concepts will be given together with the practical applications and considerations of the methods.

Even more important, though, is the understanding and interpretation of the computed models. As is methods for outlier detection and model validation. Computer exercises and cases will be performed applying user-friendly software. A thorough introduction to the software will be given.

Course content:

• Introduction to Multivariate Data Analysis
• Principal Component Analysis (PCA)
• Pre-processing
• Outlier detection
• Partial Least Squares Regression (PLSR)
• Validation
• Variable selection

Formel requirements

Basic statistical knowledge.

Learning outcome

Knowledge:
• Describe chemometric methods for multivariate data analysis (exploration and regression)
• Describe techniques for data pre-preprocessing
• Describe techniques for outlier detection
• Describe method validation principles
• Understand the basics of the algorithms behind the PCA and PLS
• Understand the math of data pre-processing

Skills:
• Apply theory on real life data analytical cases
• Apply commercial software for data analysis
• Interpret multivariate models (both exploratory and regression)

Competences:
• Discuss and respond to univariate versus multivariate data analytical methodology in problem solving in society

Target group

PhD students from any scientific field that gather data with several samples (+10) and many variables (+10).

Teaching and learning methods

There will be a mixture of several different teaching methodologies:
– Lectures (most also available as videos)
– Exercises + Walkthrough
– Short cases + Fish tank
– Day cases + Debriefing sessions
– Visual examples

Remarks

Responsible for scientific course content: Åsmund Rinnan, aar@food.ku.dk.

Collaborating departments at University of Copenhagen: PLEN and CHEM.

Disclaimer:
DDSA has explicit permission from Arcanic and the owners of the https://phdcourses.dk/ website to display the courses on ddsa.dk.