qiannanduan/A-color-spectral-machine-learning-path-for-analysis-of-five-mixed-amino-acids

Determination of mixed amino acids (AAs) often follows many time-consuming and costly operations. With development of artificial intelligence, some machine learning (ML) algorithms contained huge neural networks have shown their advantages in complex systems perception, which offer a new way for AAs analysis. Although various algorithms made success in many fields, the implementation of this idea also was limited due to the lack of relevant large data. In this work, we presented a smart ML strategy based on high-throughput experiments to measure AA concentrations in multi-component samples; and created a new data from, color spectral images, for recording spectral information of matters. The results showed that a well-trained model could predict multiple AAs at the same time, suggesting its value in facilitating quantitative analysis of mixed systems.

NOASSERTION