Physics
Scientific paper
Jun 1998
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=1998aipc..430..237b&link_type=abstract
The eleventh international conference on fourier transform spectroscopy. AIP Conference Proceedings, Volume 430, pp. 237-240 (1
Physics
1
Spectral Absorption By Atmospheric Gases, Effects Of Clouds And Water, Ice Crystal Phenomena
Scientific paper
Open-path Fourier transform infrared (OP/FT-IR) spectrometry is susceptible to interferences, especially those due to atmospheric water and carbon dioxide. To overcome these interferences, multivariate techniques such as classical least squares (CLS) and partial least squares (PLS) regression are used. The CLS method is very sensitive to variations in the background and the presence of lines due to water and carbon dioxide in the spectrum, and usually requires calculations to be made over small spectral windows to minimize these effects. The windows are usually chosen by the user and can reflect the user's bias. We have used a heuristic program called a genetic algorithm to select spectral windows that are independent of user bias and often yield results 2-4 times better than the generally accepted method of spectral window selection. We have also investigated the use of spectral window selection by a genetic algorithm for PLS using much larger spectral windows and found it to be advantageous.
Berry James R.
Griffiths Peter R.
No associations
LandOfFree
Genetic algorithms used for spectral window selection in open-path fourier transform infrared spectrometry does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Genetic algorithms used for spectral window selection in open-path fourier transform infrared spectrometry, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Genetic algorithms used for spectral window selection in open-path fourier transform infrared spectrometry will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1330425