Advanced

Analysis of neural networks in terms of domain functions

vanderZwaag, B J; Spaanenburg, Lambert LU and Slump, C H (2002) 3rd IEEE Benelux Signal Processing Symposium (SPS) In 3rd IEEE Signal Processing Symposium SPS 02 p.237-240
Abstract
Despite their success-story, artificial neural networks have one major disadvantage compared to other techniques: the inability to explain comprehensively how a trained neural network reaches its output; neural networks are not only (incorrectly) seen as a “magic tool” but possibly even more as a mysterious “black box.” Although much research has already been done to “open the box,”

there is a notable hiatus in known publications on analysis of neural networks. So far, mainly sensitivity analysis and rule extraction methods have been used to analyze neural networks. However, these can only be applied in a limited subset of the problem domains

where neural network solutions are encountered.

In this paper we... (More)
Despite their success-story, artificial neural networks have one major disadvantage compared to other techniques: the inability to explain comprehensively how a trained neural network reaches its output; neural networks are not only (incorrectly) seen as a “magic tool” but possibly even more as a mysterious “black box.” Although much research has already been done to “open the box,”

there is a notable hiatus in known publications on analysis of neural networks. So far, mainly sensitivity analysis and rule extraction methods have been used to analyze neural networks. However, these can only be applied in a limited subset of the problem domains

where neural network solutions are encountered.

In this paper we propose a wider applicable method which, for a given problem domain, involves identifying basic functions with which users in that domain are already familiar, and describing trained neural networks, or parts thereof, in terms of those basic functions. This will provide a comprehensible description of the

neural network’s function and, depending on the chosen base functions, it may also provide an insight into the neural network’s inner “reasoning.” It could further be used to optimize neural network systems. An analysis in terms of base functions may even make clear how to (re)construct a superior system using those base functions,

thus using the neural network as a construction advisor. (Less)
Please use this url to cite or link to this publication:
author
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
in
3rd IEEE Signal Processing Symposium SPS 02
pages
237 - 240
conference name
3rd IEEE Benelux Signal Processing Symposium (SPS)
language
English
LU publication?
no
id
ed2489bc-3992-4d6e-9bda-119e7a108fd2 (old id 603917)
date added to LUP
2007-12-03 11:28:12
date last changed
2016-06-29 09:01:02
@misc{ed2489bc-3992-4d6e-9bda-119e7a108fd2,
  abstract     = {Despite their success-story, artificial neural networks have one major disadvantage compared to other techniques: the inability to explain comprehensively how a trained neural network reaches its output; neural networks are not only (incorrectly) seen as a “magic tool” but possibly even more as a mysterious “black box.” Although much research has already been done to “open the box,”<br/><br>
there is a notable hiatus in known publications on analysis of neural networks. So far, mainly sensitivity analysis and rule extraction methods have been used to analyze neural networks. However, these can only be applied in a limited subset of the problem domains<br/><br>
where neural network solutions are encountered.<br/><br>
In this paper we propose a wider applicable method which, for a given problem domain, involves identifying basic functions with which users in that domain are already familiar, and describing trained neural networks, or parts thereof, in terms of those basic functions. This will provide a comprehensible description of the<br/><br>
neural network’s function and, depending on the chosen base functions, it may also provide an insight into the neural network’s inner “reasoning.” It could further be used to optimize neural network systems. An analysis in terms of base functions may even make clear how to (re)construct a superior system using those base functions,<br/><br>
thus using the neural network as a construction advisor.},
  author       = {vanderZwaag, B J and Spaanenburg, Lambert and Slump, C H},
  language     = {eng},
  pages        = {237--240},
  series       = {3rd IEEE Signal Processing Symposium SPS 02},
  title        = {Analysis of neural networks in terms of domain functions},
  year         = {2002},
}