Validation of topology evolution in artificial neural networks
(2014) FYTK01 20122Computational Biology and Biological Physics - Undergoing reorganization
- Abstract
- There is a problem with overtraining when applying Artificial Neural Networks to a classification problem. One factor that strongly influences the overtraining is the network topology. Understanding the role of the network topology is therefore important to enable the creation of more efficient Artificial Neural Networks.
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/4360879
- author
- Albert Florea, George LU
- supervisor
-
- Patrik Edén LU
- organization
- course
- FYTK01 20122
- year
- 2014
- type
- M2 - Bachelor Degree
- subject
- keywords
- TWEANN, overtraining, ANN, MLP, topology, topology evolution, classification problem, Pima Indian dataset
- language
- English
- id
- 4360879
- date added to LUP
- 2014-03-20 15:59:45
- date last changed
- 2020-01-24 13:11:46
@misc{4360879,
abstract = {{There is a problem with overtraining when applying Artificial Neural Networks to a classification problem. One factor that strongly influences the overtraining is the network topology. Understanding the role of the network topology is therefore important to enable the creation of more efficient Artificial Neural Networks.}},
author = {{Albert Florea, George}},
language = {{eng}},
note = {{Student Paper}},
title = {{Validation of topology evolution in artificial neural networks}},
year = {{2014}},
}