In:
Applied Mechanics and Materials, Trans Tech Publications, Ltd., Vol. 865 ( 2017-6), p. 630-635
Abstract:
From the research of fluorescence spectra of lactobacillus acidophilus, streptococcus mutans and lactobacillus bulgaricus, it was indicated that under the stimulation of UV light, the prbiotic solution would emit fluorescence. Under the stimulation of optimal excitation wavelength of 290nm, the fluorescence peak fell in the range of 300-650nm. 150 groups of spectral data measured by wavelet transform were used for compression, after which each group of data reduced from 1341 points originally to 168 points, which not only retained the features of the original spectrum, but also increased the processing speed of neural networks. Radial basis function neural network method was used for compressed data research, 40 groups of experimental data were trained for each strain, based on which 30 groups of unknown data were identified. Results showed that after training, the radial basis function neural network could accurately predict unknown strains.
Type of Medium:
Online Resource
ISSN:
1662-7482
DOI:
10.4028/www.scientific.net/AMM.865
DOI:
10.4028/www.scientific.net/AMM.865.630
Language:
Unknown
Publisher:
Trans Tech Publications, Ltd.
Publication Date:
2017
detail.hit.zdb_id:
2251882-4
Permalink