Human body odor discrimination by GC-MS spectra data mining

摘要: The present study explores individual identity perception by analyzing the chemical peak information in gas chromatography-mass spectrometry (GC-MS) spectra of body odor samples with standard data mining approaches. Mainly, principal component analysis (PCA) method is chosen for visual discrimination feature space. PCA combination support vector machine (SVM) used quantitative recognition. GC-MS characterization confirms composition numerous species (aldehydes, acids, ketones, esters, sulfides etc.) samples. from armpit and neck three people (from dissimilar age groups) at two different sampling times (0 h 4 h) were recorded experiment. A few blank (non-body odor) also characterized included as references further methods. efficiency (both qualitative quantitative) odors was evaluated (i) variables (the area, height ratio area to height); (ii) h); (iii) parts armpit). best has been achieved using a variable time h. This result established class separability measures calculated (PC) scores SVM classification outcomes (86%).