Discriminability graphic representation

Source discrimination was accomplished by examining a series of two- and three-dimensional plots of the obsidian source data. Discovery of graphical representations which show the clearest picture of inter-source versus intra-source variation, makes possible source discrimination with a high degree of confidence. The greater the number of elements that one can use to reinforce the observed discrimination the smaller becomes the chances for misassignment of artifacts when compared to the obsidian source database. 

Fig. 5-23. Graphical representation of fifteen objects (samples) from the interlaboratory comparison in the plane of the first two discriminant functions (seven original variables)...

Nandy, A. (1996b) Two-dimensional graphical representation of DNA sequences and intron-exon discrimination in intron-rich sequences. Comput. Appl. Biosci., 12, 55-62. 

Fig. 1. Graphical representation of discriminant function D1 derived from the che-mometric analysis of mass spectral signatures of 30 microlayer and bulk seawater film extracts (FI). Intensities at each m/z represent the loadings for that particular m/z variable used to calculate the D1 discriminant function score. Prominent m/z variables include those representative of fatty acids, acyl lipids, sterols, poloxy-mers and humic compounds...

For the PCA and PLS-DA, sparse analyses perform a selection from automatic variables. More recently, more complex methods of automatic learning from data mining have been applied to metabolomic data. Decision trees aid the automatic selection of discriminant variables, supply a simple representation of the decision model (the tree) and constitute an exploratory technique to understand complex metabolic profiles. The artificial neuron network was successfully used to classify chemical profiles and is becoming one of the most popular methods for understanding patterns. Data visualization and interactivity are now used to visualize metabolomic data in order to facilitate the interpretation of complex data-sets. XCMS online [GOW 14] offers cloud-plots, PCA and interactive heatmaps (i.e. the heatmaps are graphical representations of correlation matrices). These two types of visualization help the user personalize the display and easily select the most interesting compounds. 

FIGURE 21 Example of application of the potential function method for the discrimination of two classes in the univariate case graphical representation of (A) the cumulative potential and of (B) the posterior probability of class belonging for the two categories as a function of the value of the independent variable x. 

---