e estimated peaks with their spectral analyser values for the milk

milk data will be discussed latter.

The estimated peaks for the milk data. The texts printed in the plot represent the

alyser values (m/z) of the merged or the unique and significant peaks.

e fitness of a BWH model

mulated data set, it is possible to examine whether the discovered

and unique) peaks are correct. The following equation can be

etect whether the discovered peak locations (the spectra analyser

match the true peak locations, where ݌ stands for the ith location

analyser value) of a merged or unique peak across replicated

ݕ stands for the true peak location, y stands for a vector of all true

ations and ߜܡ stands for the difference between consecutive peak

The value of ߝ was 0.054 in a BWH model constructed for the

d data shown in Figure 5.5.

ߝൌ

1

minሺߜܡሻඩ1

ܰ෍ሺ݌െݕ

௜ୀଵ

(5.32)

gning peaks for replicated spectra

ementioned was used to estimate a baseline for a single spectrum.

a set composed of multiple spectra, an update is required. Suppose