Neural Networks and their Diagnostic Applications

C. M. Bishop

Research output: Contribution to journalArticlepeer-review


Neural network techniques offer a wide range of new opportunities for the analysis of data from plasma diagnostics. In particular, the class of neural network known as the multi-layer perceptron provides a general purpose approach to nonlinear data transformation between multidimensional spaces. In this paper, we outline the principles of the multi-layer perceptron, and illustrate its application to plasma diagnostics using two examples. The first of these concerns the extraction of line shape parameters from spectral data, and offers considerable improvements in speed compared with conventional approaches. The second application involves deconvolution of line integral data from a multichannel interferometer allowing the extraction of more detailed density provides than obtained by conventional Abel inversion.

Original languageEnglish
Pages (from-to)4772–4774
Number of pages3
JournalReview of Scientific Instruments
Issue number10
Publication statusPublished - 1 Jan 1992


Dive into the research topics of 'Neural Networks and their Diagnostic Applications'. Together they form a unique fingerprint.

Cite this