Find in Library

Possible use of compound-specific isotope analysis (GC-C-IRMS) was assessed using 255 meat samples from 24 countries. Upon results of isotope ratio (δ13C) determination in amino acids after gas chromatographic separation samples were classified using Support Vector Machines and linear discriminate analysis. Prognosis accuracy was 65-75%. Despite the fact that the use of compound analysis only provides low identification accuracy there's a possibility to considerably increase reliability of classification in combination with other data by means of extending dimension by the number of variables.