23rd Jun, 2008

CAC-2008 in Montpellier, France

The Eleventh Conference on Chemometrics in Analytical Chemistry, CAC-2008, begins next week in Montpellier, France. The conference runs from June 30 through July 4.

All indications are that it will be a great conference. The organizers say that attendance is will be close to 350, which must be a record for CAC.

Eigenvector will be there, of course. Our Jeremy Shaver will present Automatic Sample Weighting for Inferential Modeling of Historical In-Control Process Data, which is concerned with the problem of developing calibration models from data where the bulk of the samples are tightly clustered, with only a few samples exhibiting significant variation.

I’ll be there as well, presenting Tools for Multivariate Calibration Robustness Testing with Observations on Effects of Data Preprocessing. We all want calibration models that are robust, and thus, have good longevity. But how do you tell how brittle a model is? This talk demonstrates some tools for assessing model performance in the face of changes in the samples and instruments.

I’m also presenting a poster, Properties of PLS, and Differences between NIPALS and Lanczos Bidiagonalization. I’ll write about this a little more in my next post, but suffice it to say that there is a bit of controversy of late about various algorithms for Partial Least Squares Regression and the residuals they generate.

Eigenvector is of course proud to be a sponsor of CAC. We are sponsoring the Best Poster Contest, and will present the winner with $500USD (about 322€ today). I personally really like poster sessions. Its a great time to really talk with people about their research, and its generally much more of an exchange of scientific ideas than a talk, which are primarily one-way communications.

So, if you are going to CAC, look us up. Jeremy and I are always happy to answer questions about our products and services, and are always looking for user input on features for PLS_Toolbox, Solo, etc.

See you at CAC!


Leave a response

You must be logged in to post a comment.