|
|
|
The cover displays the results of a clustering analysis of a large database of mass spectra. Each spectrum consists of approximately 150 values. The number of values was reduced to five summary values through the use of principal components. A k-means clustering procedure was then performed on these values. The figure displays all pairwise scatter plots of the five summary values with the color indicating cluster membership. The analysis will aid in the matching of spectra of unidentified compounds to a compounds in the database. 
Date created: 7/20/2001 |