• English
    • Norsk
  • English 
    • English
    • Norsk
  • Administration
View Item 
  •   Home
  • Det matematisk-naturvitenskapelige fakultet
  • Institutt for informatikk
  • Institutt for informatikk
  • View Item
  •   Home
  • Det matematisk-naturvitenskapelige fakultet
  • Institutt for informatikk
  • Institutt for informatikk
  • View Item
JavaScript is disabled for your browser. Some features of this site may not work without it.

A Neuro-Evolutionary scheme for preserving and engaging in multiple, unrelated behaviors in artificial neural networks.

Bergestad, Stig
Master thesis
View/Open
The file is restricted (More info)
Bergestad.pdf (1.328Mb)
Year
2009
Permanent link
http://urn.nb.no/URN:NBN:no-22685

Metadata
Show metadata
Appears in the following Collection
  • Institutt for informatikk [4045]
Abstract
Neural networks suffer from interference when they are evolved for different and unrelated tasks.

This interference causes the networks to forget previously learned behaviours and worsens the

more the network is trained in the new task, preventing networks being able to engage in such

different unrelated tasks.

This report proposes that a cause for this may be that the regular ANN architecture is too

simple and that some additional enhancements may be needed. It is then proposed that modularity

could be used for such purpose, protecting network structures responsible for particular behaviours.

Further, it is proposed a selection scheme intended to handle the selection of modules based on

input to the primary network.

This module-with-selection scheme is tested and shows that modularity provide some

benefits such as simplifying the search space and also enabling symmetric and repeating structures

through module connectivity and reuse. The selection mechanism is also shown to be evolvable,

something not readily apparent.

It is further suggested that this selection mechanism can in the future enable networks to

operate in fractured domains. Also the scheme has support for module within module structures,

enabling a bottom-up approach to constructing networks through incremental evolution, and with

the modules acting as safeguards against interference as the network becomes more complex.
 
Responsible for this website 
University of Oslo Library


Contact Us 
duo-hjelp@ub.uio.no


Privacy policy
 

 

For students / employeesSubmit master thesisAccess to restricted material

Browse

All of DUOCommunities & CollectionsBy Issue DateAuthorsTitlesThis CollectionBy Issue DateAuthorsTitles

For library staff

Login
RSS Feeds
 
Responsible for this website 
University of Oslo Library


Contact Us 
duo-hjelp@ub.uio.no


Privacy policy