TR available: Fourier Analysis and Filtering of a Single Hidden Layer Perceptron

[payman@uw-isdl.ee.washington.edu]

 FTP-host: archive.cis.ohio-state.edu
 FTP-file: pub/neuroprose/marks.fourier.ps.Z


 The following paper is now available from the neuroprose repository:

		 Fourier Analysis and Filtering of 
		 a Single Hidden Layer Perceptron

               Robert J. Marks II & Payman Arabshahi
                
	       Department of Electrical Engineering
                  University of Washington FT-10
                       Seattle, WA 98195 USA


 This  is  an  invited  paper  to appear  in the Proceedings of the 
 International Conference on Artificial Neural Networks (IEEE/ENNS)
 Sorrento, Italy, May 1994.

		          
		 	   Abstract

 We show that the Fourier transform of the linear output of a single
 hidden  layer  perceptron consists  of  a  multitude of line masses
 passing through the origin.  Each  line  corresponds  to one of the
 hidden neurons and its slope is  determined by that neuron's weight
 vector. We also show that convolving the output of the network with  
 a function can  be achieved simply  by  modifying  the shape of the 
 sigmoidal nonlinearities in the hidden layer.



 To retrieve the file:

 unix> ftp archive.cis.ohio-state.edu
 Name: anonymous
 Password: your email address
 ftp> cd pub/neuroprose
 ftp> binary
 ftp> get marks.fourier.ps.Z
 ftp> bye

 unix> uncompress marks.fourier.ps.Z
 unix> lpr marks.fourier.ps