Affordable Access

Publisher Website

Study on SVM Calibration Model Parameter for Mixed Gas

Elsevier Ltd
Publication Date
DOI: 10.1016/j.proeng.2011.08.682
  • Support Vector Machine
  • Infrared Spectra
  • Calibration Model
  • Pattern Recognition
  • Physics


Abstract The calibration model of infrared spectra analysis is established combined with support vector machine, a new information processing method. As the model parameters have an impact on the analysis, selection of SVM calibration model parameters is researched through experimental studies. Beginning with the study of SVM calibration model parameters and spectra data sample parameters, the effect on the analysis results by the parameters like types of SVM calibration model, spectrometer scanning interval, spectra analysis band, kernel function, penalty factor C etc is researched after the infrared (IR) spectra data is preprocessed by the use of normalization-expansion method. The experimental results show that as the SVM calibration model is determined, a reasonable selection of parameters can improve the accuracy of the analysis results, and it has a practical application value.

There are no comments yet on this publication. Be the first to share your thoughts.