Methods for data processing and classification for mass spectra of exhaled gases using discriminant analysis

Currently, there are methods for estimation of human health condition based on the composition of the exhaled gas. At health surveys such methods are appears to be applicable for automatic classification of people's health groups. In this paper, the method of linear discriminant analysis is used for the classification of people's health. Discriminant analysis is a branch of multivariate statistical analysis, which includes methods for classification of multidimensional observations on the principle of maximum similarity of the analyzed observations with observations assimilated certain classes of learning outcomes. It is shown that the amplitudes of the lines of the mass spectrum can serve as values of discriminant features for classification. Examples of the testing of the proposed methods are provided.