We have shown that neural networks are capable of accurately identifying the Raman spectra of aqueous solutions of small-molecule neurotransmitters. It was found that the networks performed optimally when the ratio of the number of hidden nodes to the number of input nodes was 0.16, that network accuracy increased with the number of input layer nodes, and that input features influenced the abilities of networks to discriminate or generalize between spectra. Furthermore, networks employing sine transfer functions for their hidden layers trained faster and were better at discriminating between closely related spectra, but they were less tolerant of spectral distortions than the networks using sigmoid transfer functions. The latter type of network produced superior results where generalization between spectra was required.
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.
Contact your librarian or system administrator
Login to access OSA Member Subscription