We have been able to simulate the anomeric region of the 1H-NMR spectra of xyloglucans as a model to overcome the difficulty of obtaining sufficient spectra to train a neural network search engine. Such simulated spectra are particularly useful in situations where insufficient real spectra are available to train a neural network search engine. In such cases, if certain peak information about the anomeric region of spectra is available for all compounds in a particular database, we can then simulate this region of the spectra and train the neural networks with these simulated spectra.
Showing the most recent 10 out of 245 publications