Quantitative Spectroscopic Calibration
暂无分享,去创建一个
Spectroscopic calibration is the process of using mathematical calculations to relate the measured spectroscopic properties of a set of samples to the chemical composition (and sometimes other properties) of those samples. Any of several algorithms are employed to perform these calculations. Regardless of the algorithms used, however, successful execution of this process depends on properly performing several tasks before and after the main calibration algorithm is applied: selecting a suitable set of calibration samples to ‘train’ the system, using the correct algorithm to match the characteristics of the measured data, and properly evaluating and validating the calibration model developed. In this article, these various aspects of calibrating spectroscopic instruments to perform quantitative analysis are discussed. The article ends with an overview of the various algorithms that are available to perform the calibration process.
[1] Tormod Næs,et al. Artificial Neural Networks in Multivariate Calibration , 1993 .
[2] Riccardo Leardi,et al. Genetic Algorithms as a Tool for Wavelength Selection in Multivariate Calibration , 1995 .