HOME
13th ANISG CONFERENCE AND SHORT COURSE
Invitation and call for papers
 
Conference format and programme
 
Invited guest speakers
  
Short course
 
The Kaji Competition
Trade display and sponsorship
 
Lynsey Welsh Award
 
Venue and accommodation
 
Social programme
Registration - short course and/or conference
Student awards
 
Oral and poster presentations

OVERTONES
Latest issue (September 2007)
 
Back issues

LINKS
International Council for Near Infrared Spectroscopy
 
NIR Publications
 
Pittsburgh Conference 2008
 
Chambersburg Conference 2008
Eastern Analytical Symposium 2008
14th ICNIRS Conference (NIR-2009)


The Kaji Competition

You are invited to enter a chemometrics competition, which will again feature at the 13th ANISG conference at Hamilton.

An aboriginal word, “Kaji”, meaning to seek, has been chosen as the name for this competition.  A similar competition has been a feature of the International Diffuse Reflectance Conference at Chambersburg, USA for some years. 

A set of NIR spectral data will be available for downloading from the ANISG website and contestants will be asked to find and explain a "best" chemometric model to robustly predict samples of the same type.  A panel will select the three "best" entries based on the predicted results and spectroscopic explanation of the products and attributes of interest.

Welcome to the Kaji Competition for the 2008 ANISG Conference.  We apologise for the delay in making the information and data available.

There are two data files provided – a calibration file containing spectra and reference values (157 samples), and a validation file containing spectra only (30 samples).

The files were created in ISI format and the spectra collected on a Foss 6500 instrument.  We have also converted the files to ASCII and JCAMP format.

The reference data are designated as attributes A, B and C.  This is a “real world” data set, and you may notice an occasional “bad” spectrum.

The idea is to create calibration equations for all three attributes using the calibration file, and then predict these attributes on the validation file (spectra only).  Calibration statistics should be provided.  The judges have access to the reference data on the validation file, and will assess each entrant’s predicted values against the reference values.  Any comments on the spectra, the reference values and the structure of the sample population will be welcome.

The file names are as follows:

Calibration file:     
                              K08CAL.CAL (ISI format)
                              K08CAL.DAT (ASCII format)
                              K08CAL.JCM (JCAMP format)
                              K08CAL.CFL

Validation file:   
                              K08VAL.NIR (ISI format)
                              K08VAL.DAT (ASCII format)
                              K08VAL.JCM (JCAMP format)
                              K08VAL.CFL

Good luck!


ENTRIES MUST BE E-MAILED TO PETER FLINN BY NO LATER THAN MONDAY 7TH APRIL.