Basic Chemometrics and a little machine learning

Background: Chemometrics (or multivariate data analysis) may be used to solve problems involving large amounts of data. This is relevant within fields such as development, research, process monitoring and control, and laboratory analysis. In these fields, the use of single variables is often inadequate to describe, differentiate or classify objects/samples. Looking at more variables at a time ensures that interactions, patterns and correlations are taken into consideration. Combined with superior data visualization, chemometrics is a needed tool for proper data analysis.

As participant you will be introduced to the multivariate way of thinking and learn how to explore your data properly and how to set up a multivariate calibration/regression model. The course is a mixture of lectures and exercises. In the exercises, you will use the chemometrics tools. You will learn to navigate through the raw data, develop models and visualize the models. During this course you will not be able to work with your own data.

Audience: The course is intended for people handling problems where chemometrics may be applied or people who have a general interest in learning more about chemometrics and its applications. Some mathematical and statistical expressions will be used in the course and a variety of data (e.g. sensory and spectroscopic data) will be used as examples.

Software: PLS_Toolbox or SOLO will be used during exercises and has to be installed on your own laptop. Instructions for obtaining a demo version will be sent before the course.

Teachers: Rasmus Bro

Location: University of Copenhagen, Frederiksberg Campus

The course is taking place from 9 AM to 5 PM both days. Lunch and coffee will be included. If you have special dietary needs, please let us know by enrollment. Lectures and notes are in English.

Enroll to Rasmus Bro (rb@food.ku.dk) March 18th at the latest. Cancellations must be made no later than four days in advance.

Price: 7000 DKK

See also Two day course in Design of Experiments