Comments on “parametric analysis of histograms measured in flow cytometry” by R.C. Mann et al

✍ Scribed by F. G. Boese

• Publisher: John Wiley and Sons
• Year: 1986
• Tongue: English
• Weight: 191 KB
• Volume: 7
• Category: Article
• ISSN: 0196-4763
• DOI: 10.1002/cyto.990070217

✦ Synopsis

In a paper by R.L. Mann et al. (6), the authors dealt with a principal step in processing spectroscopic histograms. They considered the so-called mixture problem for the univariate normal case using the statistical approach.

Given a histogram where the XkcR are the variate values defining the histogram classes and the h&+ the corresponding frequencies, suppose that H stems from a sample that was drawn from a parent distribution with the mixture density beginning with an appropriate chosen vector y(O'. The authors stated correctly that the iteration 4 will converge if for any matrix norm 11 * 1) (compatible with the vector norm used in Rd) the Jacobian 9' of @ obeys 11 @'(y*) 11 < x , h < 1