Journal of
Systemics, Cybernetics and Informatics
 



ISSN: 1690-4524 (Online)


Indexed by
EBSCO, Cabell, DOAJ (Directory of Open Access Journals)Benefits of supplying DOAJ with metadata:
  • DOAJ's statistics show more than 900 000 page views and 300 000 unique visitors a month to DOAJ from all over the world.
  • Many aggregators, databases, libraries, publishers and search portals collect our free metadata and include it in their products. Examples are Scopus, Serial Solutions and EBSCO.
  • DOAJ is OAI compliant and once an article is in DOAJ, it is automatically harvestable.
  • DOAJ is OpenURL compliant and once an article is in DOAJ, it is automatically linkable.
  • Over 95% of the DOAJ Publisher community said that DOAJ is important for increasing their journal's visibility.
  • DOAJ is often cited as a source of quality, open access journals in research and scholarly publishing circles.
JSCI Supplies DOAJ with Meta Data
, Academic Journals Database, and Google Scholar


Listed in
Cabell Directory of Publishing Opportunities and in Ulrich’s Periodical Directory


Re-Published in
Academia.edu
(A Community of about 40.000.000 Academics)


Honorary Editorial Advisory Board's Chair
William Lesso (1931-2015)

Editor-in-Chief
Nagib C. Callaos


Sponsored by
The International Institute of
Informatics and Systemics

www.iiis.org
 

Editorial Advisory Board

Editors

Journal's Reviewers
 

Description and Aims

Submission of Articles

Areas and Subareas

Information to Contributors

Editorial Peer Review Methodology

Integrating Reviewing Processes


Detection of Minimal Set of Trips Causing the Necessity to Use Extra Vehicle for Vehicle Scheduling Problem
Katerina Pastircáková, Jaromír Šulc
(pages: 1-4)

Key Factors in the Success of Self - Directed Learning of Military Personnel - Taking Smartphone as an Example
Yen-Hsi Lo, Yen-Fen Lo, Po-Yun Chiang, Jung Hsiao
(pages: 5-8)

Machine Learning Based IP Network Traffic Classification Using Feature Significance Analysis
Te-Shun Chou, John Pickard, Ciprian Popoviciu
(pages: 9-12)

The Information System for US Stock Market: Fundamental and Technical Analysis
Sergejs Hilkevics, Galina Hilkevica
(pages: 13-24)

The Impact of Environmental and Social Performance on the Market Value of Shares of Czech Joint-Stock Corporations
Alena Kocmanova, Marie Pavlakova Docekalova, Iveta Simberova
(pages: 25-31)

Play the Game! Analogue Gamification for Raising Information Security Awareness (Invited Paper)
Margit Scholl
(pages: 32-35)

Using Informatics and Technology Practices for Academic Performance Review
Kim Moorning
(pages: 36-41)

Multiple Research Perspectives as a Paradigm to Co-Create Meaningful Real-life Experiences
Jan Detand, Marina Emmanouil
(pages: 42-46)

A Methodology to Integrate Regulatory Expertise, Research and Education to Accelerate Biomedical Device Translation
Diana Easton
(pages: 47-52)

Active Learning through Smart Grid Model Site in Challenge Based Learning Course
Ellen A. Kalinga, Kwame S. Ibwe, Nerey H. Mvungi, Hannu Tenhunen
(pages: 53-64)

Non-Linear Static Analysis of Masonry Buildings under Seismic Actions
Maria Luisa Beconcini, Paolo Cioni, Pietro Croce, Paolo Formichi, Filippo Landi, Caterina Mochi
(pages: 65-70)

Toward an Engaging Hands-on Environment for a Beginning Networking and Security Class
Lopamudra Roychoudhuri
(pages: 71-76)

Designing Representations, Affecting Reality: A Meta-Model Proposal to Address the Question of Design Epistemology from the Perspective of Cognitive Science
Andrea Zammataro
(pages: 77-80)

Dielectrophoretic Movement of Cell around Surface Electrodes in Flow Channel
Yusuke Takahashi, Shigehiro Hashimoto, Manabu Watanabe
(pages: 81-87)


 

Abstracts

 


ABSTRACT


Influence of the Training Methods in the Diagnosis of Multiple Sclerosis Using Radial Basis Functions Artificial Neural Networks

Ángel Gutiérrez


The data available in the average clinical study of a disease is very often small. This is one of the main obstacles in the application of neural networks to the classification of biological signals used for diagnosing diseases. A rule of thumb states that the number of parameters (weights) that can be used for training a neural network should be around 15% of the available data, to avoid overlearning. This condition puts a limit on the dimension of the input space.

Different authors have used different approaches to solve this problem, like eliminating redundancy in the data, preprocessing the data to find centers for the radial basis functions, or extracting a small number of features that were used as inputs. It is clear that the classification would be better the more features we could feed into the network.

The approach utilized in this paper is incrementing the number of training elements with randomly expanding training sets. This way the number of original signals does not constraint the dimension of the input set in the radial basis network. Then we train the network using the method that minimizes the error function using the gradient descent algorithm and the method that uses the particle swarm optimization technique.

A comparison between the two methods showed that for the same number of iterations on both methods, the particle swarm optimization was faster, it was learning to recognize only the sick people. On the other hand, the gradient method was not as good in general better at identifying those people.

Full Text