Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!ucsd!swrinde!zaphod.mps.ohio-state.edu!sunybcs!uhura.cc.rochester.edu!rochester!kodak!uupsi!sunic!liuida!isy!nicwi
From: nicwi@isy.liu.se ( Niclas Wiberg)
Newsgroups: comp.ai.neural-nets
Subject: Fault tolerance in internal representation
Keywords: fault tolerance, channel coding, internal representation
Message-ID: <1990Jul25.114212.17172@isy.liu.se>
Date: 25 Jul 90 11:42:12 GMT
Distribution: comp
Organization: Dept of EE, University of Linkoping
Lines: 16


I am studying the internal representation in multi-layer feed-forward
networks. I am specially interested in the fault tolerance capabilities
of the representation, from a channel-coding point of view. I am trying
to improve these capabilities by adding some random noise to the
internal signals during the training phase.

I wonder if anybody has done some related work, and what references
there are.

Please answer by email.

-- 
----------------------------------------------------------------------
Niclas Wiberg                         nicwi@isy.liu.se
Dept. of EE    Linkoping University   Sweden