Research in Chemometrics, Artificial Intelligence and Machine Learning at UCPH FOOD
The Data Analytics Group at The Department of Food Science at the University of Copenhagen (UCPH FOOD) has been doing research on machine learning (ML) and artificial intelligence (AI) for more than 25 years. The application of these tools in the food industry and food chemistry is called chemometrics.
We analyze food samples and get huge amounts of data that can be used for solving complex problems and uncovering completely new phenomena. The aim is not only predictive performance but also to get a better understanding of the food system. This will assist engineers and food researchers to explore and interpret complex biological systems.
At UCPH FOOD we use data analysis to measure properties and functionalities from raw materials to end-products so that we can optimize and characterize food production and end-quality. This is widely applied for raw material identification, process monitoring and optimization, food fraud and adulteration detection, gastronomy, understanding the biology of fermentation flavor research, proteomics, metabonomics and many other areas. (See also Foodomics at UCPH FOOD)
As our field is data analysis, our basic research is concerned with statistics, artificial intelligence, mathematical optimization and machine learning, but it is also tied to the physical- and chemical nature of the data observed. In short, chemometrics.
Process monitoring: We work with process analytical technology, quality by design and statistical process chemistry as well as process monitoring to improve, optimize, monitor and ensure the quality in food production. We run a specialized MSc in this area and we are directly linked to the research section developing dedicated measurement technologies (Food Analytics and Biotechnology).
Metabonomics: We work with metabonomics especially in relation to health, gastronomy and food quality to better understand e.g. physical properties of food and we are directly linked to the research section involved in data collection and experimental designs of metabonomic studies (Food Analytics and Biotechnology).
Spectroscopy: We work with many types of advanced analytical instruments such as near-infrared spectroscopy, fluorescence, nuclear magnetic resonance, chromatography etc. We develop specialized software and algorithms for extracting information from such data; software/algorithms that we freely share at http://models.life.ku.dk/algorithms.
We offer many courses from the beginning of the bachelor to advanced PhD and covering all aspects of chemometrics, machine learning, design of experiments, metabolomics, process analytical technology, quality by design and many other subjects.
You can find some of the courses here and here or use the search function at kurser.ku.dk.
We collaborate with all fields and facilities at UCPH FOOD.
We have extensive collaboration with food companies in Denmark as well as abroad. Being the largest academic chemometrics group in the world, we also have extensive collaboration across the world. We typically have 5-15 visiting researcher mainly from academia. We collaborate extensively with PLEN, NEXS, SUND, PHARMA and other internal University of Copenhagen departments. We also have solid collaborations with different departments at DTU.
Our collaboration with industry is to solve problems or challenges they have that we deem possible to solve through the use of chemometrics. This is often connected with our focus areas, such as process monitoring, metabonomics and, in particular, spectroscopy. To facilitate this, we run intensive one-and two-day courses on chemometrics, design of experiments, spectroscopy etc. targeting industry partners through the ODIN consortium. We also do industry-partnered research programmes and master projects deploying our knowledge, resources and hours on industry-driven problems.
Frans van der Berg
Morten A. Rasmussen