header image
Home arrow Tutorial arrow Variable subset selection in linear regression
Variable subset selection in linear regression PDF Print E-mail
Written by Agnar Höskuldsson, DTU   
Many companies are collecting large amount of data and frequently it is a basic issue, which part of data should be used in the analysis. For instance, companies are investing in NIR instruments for on-line control of the production. Which part of the spectra should be used in the analysis? Many companies are basing their instruments on the NIR technology. Foss-Electric (here shortened to Foss) in Denmark is selling measurement instruments based on the NIR technology for more than 200 mio euros a year. Each instrument is calibrated to measure what it is designed for. The instrument selects appropriate parts of the spectra and by using a linear regression analysis it estimates typically between three and seven chemical parameters. The company has successfully implemented the technology in many different types of instruments. An important question is: Is it possible to establish an automatic procedure that finds the appropriate parts of the spectra that should be used such that the regression analysis gives efficient results? This is the question that will be treated in the following Tutorial.
< Prev   Next >
Search website

Editorial flash
Svante Wold's New Year Editorial 2005
Another year has passed too quickly, and we now read January 2005 on our calendars. In 2004, good things happened and terrible, happy and sad, as usual. However, chemometrics does well, spreading deeper into biology where the data sets are larger than ever with thousands of variables in, among others, gene arrays, LC-MS profiles, and NIR and IR microscopy.
Read more...
News flash
Chemometrics’ Epistemology in Systems Biology

10 years of Systems biology research – current knowledge and future expectations

May 9-11 2011, Skanör-Falsterbo, Öresund, Sweden

Systems biology has now matured and includes advanced modelling of biological systems.The condensation of divergent scientific disciplines (clinicians, biologists, mathematicians, statisticians, physicists, chemists, etc.) has forced us into a new way of working and thinking – Systems biology. This provides further challenges how to create meaningful data, extract information and ultimately gain new knowledge.

Download flyer (2010-06-10)

The focus of our conference theme, Chemometrics’ Epistemology in Systems biology will be on the scientific process, from data generation to quality of gained knowledge and its presentation to a scientific audience. Epistemology means theory of knowledge and addresses the questions: What is knowledge? How is knowledge acquired? What do systems biologists know? How do we know what we know? How knowledge convince fellow scientists? These issues will be given ample time at this EUCHEM conference, which will be limited to maximum 70 participants. Each half day will include two one-hour speeches of the same topic from two speakers representing differing point of views. The suggested format will allow plenty of time for informal discussions and networking. Further, poster presenters will be given the opportunity to give a five minutes oral presentation. This gives an opportunity also for PhD students and young researchers to contribute to the symposium.

Confirmed speakers:
Prof. Jonas Bergqvist, Dep. Analytical Chemistry, Uppsala University, Sweden
Dr. Mats Sundgren, AstraZeneca R&D Mölndal, Sweden
Dr. Nicholas Waters, CEO Neurosearch Sweden
Scientific board:
Prof. Torbjörn Lundstedt, Uppsala University & AcurePharma AB, Sweden
Dr. Mats Sundgren, AstraZeneca R&D Mölndal, Sweden
Prof. Mark Stitt, Max Planck Institute of Molecular Plant Physiology, Germany

Organization committee:
Johan Trygg (Umeå University, Umeå, Sweden)
Jenny Forshed (Karolinska Institute, Stockholm, Sweden)
Johan Gottfries (Gottfries Medicinal, Gothenburg, Sweden)
Ing-Marie Olsson (Umetrics AB, Malmö, Sweden)

 

 

 

 

 

 

 

 

 

 

 

 
Tutorial flash
Statistical Molecular Design, part 2
In this sequel, we will address SMD from another viewpoint, that of Quantitative Structure-Activity Relationships (QSAR), in which SMD is instead used to select a sub-set of compounds drawn from a larger inventory of chemicals.
 
Read more...