Aarhus University Seal / Aarhus Universitets segl

Introduction to multivariate data analysis (chemometrics) (2019)

 

Name of course:
Introduction to multivariate data analysis (chemometrics)

ECTS credits:
3 ECTS

Course parameters:
Language: English
Level of course: PhD course
Time of year: April/May 2019
No. of contact hours/hours in total incl. preparation, assignment(s) or the like:

  • 3 days of lectures and exercises (24 h)
  • 1 day of working with own data (8 h)
  • Preparation by reading selected book chapters and articles (20 h)
  • Writing report and prepare presentation (20 h)
  • Follow up 1 day presenting data analysis (8 h)

Capacity limits: Minimum 10 and maximum 25 participants

Objectives of the course:
The purpose of the course is to give an introduction to some of the common methods in multivariate data analysis, and give the students tools and knowledge to understand and perform PCA and PLS data analysis on their own data.

Learning outcomes and competences:
At the end of the course, the student should be able to:

  • Arrange data in a matrix appropriate for PCA and PLS
  • Apply PCA (exploration) and PLS (regression) on new data and analyze the results
  • Compare and contrast the methods for a given data analysis situation considering the benefits and the pitfalls of the methods
  • Apply the most common standardization methods appropriately
  • Examine relevant plots for outliers in PCA and PLS and thereby classify severe outliers, consider borderline cases and argue for the classification
  • Apply appropriate validation of PLS models and consider the number of PLS components
  • Outline the most common preprocessing methods
  • Outline classification methods such as SIMCA and discriminant PLS
  • Interpret PCA and PLS models described in scientific literature and describe your own results in a scientific way
  • Critically evaluate other students work based on model parameters and knowledge of model characteristics

Compulsory programme:
Attendance for a minimum of 80% of the theoretical and practical lessons is required to obtain the course diploma. Approved report.

Course contents:
Multivariate data analysis (chemometrics) can be used to solve problems involving large amounts of multivariate data generated by e.g. spectroscopy, chromatography or time series of many variables. In chemometrics informative patterns are found and interpreted instead of looking at classical, and often inadequate, univariate measures. Chemometrics is widely used in science and in scientific papers. It is important to know what features to use, how to use them correctly and how to interpret plots. Chemometrics include hypothesis generating methods, but can also be used for classification and prediction.

The course will give a thorough introduction to the chemometric methods, Principal Component Analysis (PCA) and Partial Least Squares (PLS) regression, including common data pre-processing.

Some mathematical and statistical expressions will be used in the course and a variety of data (e.g. chemical, sensory and spectroscopic data) will be used as examples.

Prerequisites:
Enrolled in a science-based PhD programme. Master’s students can participate as a part their Master’s project in agreement with the supervisor.

Name of lecturer:
Assistant Professor Ulrik Kræmer Sundekilde, Department of Food Science, Aarhus University, Denmark

Type of course/teaching methods:
Lectures, computer exercises, data analysis of your own data, writing and presenting report.

Literature:
Selected chapters and papers will be announced later. You will receive a folder with information and printouts from the lectures at the beginning of the course.

Course homepage:
None

Course assessment:
Passed/not passed assessment based on written report, presentation and discussion of results considering the learning outcomes.

Provider:
Department of Food Science, Aarhus University

Special comments on this course:
Course fee: 1750 DKK for PhD and Master’s students and AU-FOOD staff (covers 1 year software license and coffee/bread/fruit in the morning and afternoon), 5000 DKK for others.

The course is organized in combination with the PhD course ‘Introduction to metabolomics’ and it is possible to follow both courses, although the courses can be taken individually, if necessary. A discount is available if you follow both courses - see below.

Time:
Chemometrics:

  • Monday 29 April: Lectures
  • Tuesday 30 April: Lectures
  • Monday 6 May: Lectures
  • Tuesday 7 May: Workshop: Working with own data
  • Tuesday 14 May: Deadline for handing in report
  • Monday 20 May: Examination seminar (peer-feedback, teacher-feedback)

Metabolomics:

  • Thursday 25 April: Lectures
  • Friday 26 April: Lectures
  • Tuesday 7 May: Data preparation, workshop
  • Tuesday 14 May: Deadline for handing in report
  • Tuesday 21 May: Notice of assessment

Place:
AU Konferencecenter, Fredrik Nielsens Vej 4, 8000 Aarhus

Registration:
Deadline for registration is 1 April 2019. Information regarding admission will be sent out no later than two workdays after registration deadline.

Link to registration

The course fees are:
Metabolomics only (covering bread/coffee/fruit and folder):
PhD students and master students enrolled at Danish Universities and AU-FOOD staff: 750 DKK
Others: 3000 DKK

Chemometrics only (covering bread/coffee/fruit, folder and LatentiX license):
PhD students and master students enrolled at Danish Universities and AU-FOOD staff: 1750 DKK
Others: 5000 DKK

Metabolomics and Chemometrics (covering bread/coffee/fruit, folder and LatentiX license):
PhD students and master students enrolled at Danish Universities and AU-FOOD staff: 2000 DKK
Others: 7500 DKK

 

If you have any questions, please contact Ulrik Sundekilde, e-mail: uksundekilde@food.au.dk.

15374 / i43