Application of genetic algorithms in fundamental study of non-invasive measurement of human blood glucose concentration with near infrared spectroscopy

Genetic algorithms is an effective method in wavelength selection applied in building multivariate calibration model based on partial least squares regression. If genetic algorithm is run repeatedly as a block, the optimal solution is obtained faster, the numbers of wavelengths used to build calibration model is further reduced, the prediction precision is further improved. This method was applied to fundamental study of Non-Invasive of measurement of human blood glucose concentration with spectroscopy. The experiments tested in this method are 1 glucose in aqueous matrix, 2 glucose in aqueous matrix containing bovine serum albumin and heucoglobin and 3 human serum containing glucose. The result shows that the numbers of wavelengths for building the models can reduce by 88%, 86% and 85% respectively, while the root mean square error of prediction reduces by 56%, 64% and 63% respectively. It is instructive for the further study of the theory of non-Invasive measurement of human blood glucose with spectroscopy.